An online parameter adjustment method for a UAV system controller

Abstract

The application discloses a parameter online adjustment method of a UAV system controller, wherein the system controller comprises an obstacle avoidance controller and a target arrival controller; in the target arrival controller, a time-varying robust control barrier function comprises a predicate function corresponding to a target area and a constraint parameter determined based on a signal temporal logic task; during operation of the UAV, when it is monitored that the time-varying robust control barrier function is a non-positive number, the constraint parameter is updated according to a preset rule, and whether to continue to iteratively update the constraint parameter is selected according to a positive or negative condition of a control barrier function corresponding to the constraint parameter within a preset period. Through the parameter online adjustment method, the UAV can better meet complex requirements of the signal temporal logic task under unknown environment or limited control input conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) control, and in particular to an online parameter adjustment method for an UAV system controller. Background Art

[0002] In the field of modern control theory, control obstacle functions, as a key technology, have been widely used in the design of controllers for unmanned aerial vehicle (UAV) signal sequential logic tasks. This is an extension of the obstacle function for open-loop systems. Researchers have demonstrated a method that combines control obstacle functions with quadratic programming and has applied it to safety-critical systems, paving the way for subsequent research. To better meet the complex requirements of signal sequential logic tasks, a new control obstacle function design strategy has been proposed, which uses quadratic programming to generate controllers. This method has been further developed to handle risky signal sequential logic tasks faced by non-holonomic systems, exploring strategies for achieving effective control in uncertain environments.

[0003] However, these methods face challenges in managing the conflict between goal attainment and obstacle avoidance. The presence of obstacles can render the quadratic programming problem unsolvable, thus compromising control effectiveness. To overcome this challenge, researchers have introduced a novel control framework that combines an artificial potential field with a control obstacle function. This effectively resolves the conflict between goal attainment and obstacle avoidance, ensuring the solvability and performance of the control problem.

[0004] Even so, once the obstacle avoidance mechanism in the framework is activated, the reachability of the target point depends heavily on the choice of control obstacle function parameters. The predefined control obstacle functions in existing schemes often fail to meet the complex requirements of signal sequential logic tasks in unknown environments or with limited control input conditions. Summary of the Invention

[0005] The purpose of the present invention is to overcome the above-mentioned defects or problems existing in the background technology and provide a method for online adjustment of parameters of a drone system controller, through which the drone can better meet the complex requirements of signal timing logic tasks in unknown environments or limited control input conditions.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] Technical Solution 1: A method for online parameter adjustment of a drone system controller, wherein the drone completes a signal timing logic task under the control of the system controller; the system controller includes an obstacle avoidance controller generated based on an artificial potential field function and a target arrival controller generated based on a time-varying robust control obstacle function; in the target arrival controller, the time-varying robust control obstacle function includes a predicate function corresponding to a target area and constraint parameters determined based on the signal timing logic task; during the operation of the drone, the constraint parameters are updated in real time through the following steps: Step 1: Obtain the predicate function and robustness parameters corresponding to the target area; Step 2: Initialize the time parameters and constraint parameters based on the properties of the signal timing logic task currently being performed by the drone; Step 3: When the time-varying robust control obstacle function is non-positive, update the constraint parameters according to a preset update rule; Step 4: Within a preset period, determine whether the time-varying robust control obstacle function corresponding to the constraint parameter is less than or equal to 0, if so, return to step 3 to continue updating the constraint parameter, otherwise maintain the constraint parameter obtained after the last update; wherein, the time-varying robust control obstacle function is composed of In other words, in the proposition φ s The kth time-varying robust control barrier function is The predicate function corresponding to the target area is h g,s,k (x), the constraint parameter is γ s,k (t), and in, The initial state of the drone is x0=x(t0), let Then γ s,k (t) Initialize and update the constraint parameters in step 3 based on the following rules: In the above rules, ρ s,k is the robustness parameter, ρ s,k >0; is the time parameter, Represents the predicate function h corresponding to the predicate logic μ g,s,k (x) In Proposition F [a,b] ψ, Represents the predicate function h corresponding to the predicate logic μ g,s,k (x) In Proposition G [a,b] In ψ, ψ is the area of ​​the signal timing logic task currently being performed by the UAV; Indicates When the predicate function h g,s,k The supremum of (x), r s,k is the robustness coefficient for the proposition, r s,k ∈R ＞0 , R ＞0is a pre-set set of positive real numbers; Δ s,k To update the constraint parameters, Δ s,k ∈(0,γ s,k,∞ ).

[0008] Technical solution 2 based on technical solution 1: In step 2, the time parameter is initialized as follows: when φ s =φ1, t0=0; when φ s When s≥2, t0=b s-1 .

[0009] Technical solution 3 based on technical solution 2: In step 3, when the current time t is less than the time parameter and When it is less than or equal to 0, let t new =t, Update γ based on the update rule s,k,0 , juxtapose h g,s,k (x0)=h g,s,k (x), at this time And according to Calculate γ s,k (t); among them, Greater than or equal to the preset l max hour,

[0010] Technical solution 4 based on technical solution 3: In step 3, when the current time t is greater than or equal to the time parameter and When t is less than or equal to 0, let t new =t, update γ based on the update rule s,k,0 , juxtapose h g,s,k (x0)=h g,s,k (x), according to γ s,k (t) = (γ s,k,0 -γ s,k,∞ )exp(-l max (tt new ))+γ s,k,∞ Calculate γ s,k (t).

[0011] Technical solution 5 based on technical solution 4: During the update process of the constraint parameters, if If it is greater than 0, the γ obtained after the last update is kept s,k (t).

[0012] Technical solution 6 based on technical solution 1: the obstacle avoidance controller is generated based on the artificial potential field function; for each proposition φ s , its artificial potential field function is defined as Us (x) = U attr,s (x)+∑ k U rep,s,k (x); where U attr,s (x) is the gravitational potential field, For the drone in Proposition φ s The target state, π o,s,k is the set of obstacles perceived by the UAV at time t The predicate function h of the kth obstacle in o,s,k The subzero level set of (x(t)), With r o,s,k is φ s The center and radius of the kth obstacle in U rep,s,k (x) is φ s The repulsive potential field of the k-th obstacle in , The choice satisfies And function The zero-level set of k does not intersect, s ,η s,k is a positive scalar parameter; the obstacle avoidance controller u cpf,s Generated by solving the following quadratic programming problem: Among them, c s To adjust the obstacle avoidance rate parameters, c s >0; Based on the solution of the above quadratic programming problem, the obstacle avoidance controller u cpf,s In the following form:

[0013] in, To solve the above quadratic programming problem, we obtain the optimal controller; C g 、C d is the threshold of gradient and distance, C g >0,C d >0; δ is bounded perturbation, δ∈R m ; C g 、C d and δ satisfy the following conditions: ||g(x)δ||≤C δ , and the set In each time interval t∈[a s ,b s ] is not an invariant set.

[0014] Technical Solution 7 based on Technical Solution 6: The target arrival controller is generated based on the time-varying robust control obstacle function; for each proposition φ s , the target reaches the controller u cbf,s Generated by solving the following quadratic programming problem: Where Q∈R m×m is a positive definite matrix, α s :(-∞,+∞)→(-∞,+∞) is the expansion Class function, satisfying α s (0)=0 and strictly monotonically increasing, is the time-varying robust control barrier function, Where η>0.

[0015] Technical solution eight based on technical solution seven: the system controller u s Including the obstacle avoidance controller u cpf,s and the target arrival controller u cbf,s , which is of the following form: u s =λ s (x,h ob,s (x,t))u cbf,s +(1-λ s (x,h ob,s (x,t)))u cpf,s ; Among them, h ob,s (x) is the smooth lower approximation of the predicate function of all obstacles perceived by the drone, η ob >0,λ s (x,h ob,s (x)) is defined as Ensure λ s (x,h ob,s (x,t))∈[0,1],β s is the decision function λ s (x,h ob,s (x,t)) decay rate constant.

[0016] Technical Solution 9 based on Technical Solution 8: In step 4, for any k, if there is a moment In the time interval The constraint parameter γ s,k (t) No update occurs and in the time interval Internal s,k (t)>0, determine each sequential logic task φ s All completed.

[0017] From the above description of the present invention, it can be seen that compared with the prior art, the present invention has the following beneficial effects:

[0018] The present invention provides an online parameter adjustment method for a drone system controller. This method generates an obstacle avoidance controller based on an existing artificial potential field function and a target reach controller based on an existing time-varying robust control obstacle function. These two functions are combined to form a system controller. The system controller controls a drone to complete a pre-set signal sequential logic task, enabling the drone to avoid any perceived obstacle when approaching it, and to fly toward the target location to complete the target sequential logic task when moving away from it. However, during actual research, the inventors discovered that relying solely on the aforementioned system controller, when faced with unknown environments or limited control input conditions, often fails to ensure forward invariance of the corresponding control obstacle function, thereby affecting the drone's task performance. To address this issue, the inventors implemented an online adjustment of the constraint parameters in the time-varying robust control obstacle function, enabling the drone to adjust the corresponding constraint parameters in real time based on its current state information and the nature of the signal sequential logic task. This allows the drone to better complete complex signal sequential logic tasks under unknown environmental disturbances or limited control input conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flow chart of a method for online parameter adjustment provided by an embodiment of the present invention;

[0021] Figure 2 A schematic diagram of a drone experiment scenario provided by an embodiment of the present invention;

[0022] Figure 3 A trajectory diagram of a drone completing a mission according to an embodiment of the present invention;

[0023] Figure 4 A graph showing the changes in h(x) and γ(t) values ​​when the drone provided in an embodiment of the present invention completes a mission;

[0024] Figure 5 The control obstacle function of the drone provided by the embodiment of the present invention when completing the task The numerical change diagram of ;

[0025] Figure 6 A diagram showing the numerical changes in the speed control instruction u of a UAV provided in an embodiment of the present invention when completing a mission. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are preferred embodiments of the present invention and should not be regarded as excluding other embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0027] In the claims, description and drawings of the present invention, if the terms "include", "have" and their variations are used, they are intended to mean "including but not limited to".

[0028] An embodiment of the present invention provides an online parameter adjustment method for an unmanned aerial vehicle (UAV) system controller, wherein the UAV completes a signal timing logic task under the control of the system controller; the system controller includes an obstacle avoidance controller generated based on an artificial potential field function and a target arrival controller generated based on a time-varying robust control obstacle function.

[0029] Among them, in the target arrival controller, the time-varying robust control obstacle function includes the predicate function corresponding to the target area and the constraint parameters determined based on the signal timing logic task; and, referring to Figure 1 During the operation of the UAV, the constraint parameters are updated in real time through the following steps:

[0030] Step 1: Get the predicate function and robustness parameters corresponding to the target area.

[0031] Step 2: Initialize time parameters and constraint parameters based on the nature of the signal timing logic task currently being performed by the UAV;

[0032] Step 3: When the time-varying robust control barrier function is non-positive, updating the constraint parameters according to a preset update rule;

[0033] Step 4: Within a preset period, determine whether the time-varying robust control barrier function corresponding to the constraint parameter is less than or equal to 0. If so, return to step 3 to continue updating the constraint parameter. If not, maintain the constraint parameter obtained after the last update.

[0034] The following further describes the above-mentioned system controller and the updating process of the constraint parameters.

[0035] In this embodiment, R is defined as a set of real numbers, R n is an n-dimensional real vector space, R >0 ,R ≥0 Let t be the set of positive and non-negative real numbers, respectively. Let dist(x,A) be the shortest distance from vector x to set A, and traj(x) be the trajectory of vector x(t). Let the Euclidean norm be ||·||.

[0036] In this embodiment, the UAV system is defined as a nonlinear input affine model: Formula (1) Where x∈R n and Represents the system state and system control input; function f:R n →R n and g:R n →R n×m Satisfies local Lipschitz continuity; the drone state x is in a finite space The matrix function g of the UAV system satisfies all Satisfy g(x)g(x) T Zhengding.

[0037] The above-mentioned UAV system (1) is limited by the upper-level signal sequential logic task specification, and evaluates the predicate function h:R n →R to determine predicate logic Among them, the syntax of the signal sequential logic task φ is defined as: φ1 and φ2 are signal sequential logic task propositions, a, b∈R ≥0 And a≤b.

[0038] The semantic definition of the above signal sequential logic task proposition φ is:

[0039]

[0040] Specifically, in this embodiment, the target arrival signal timing logic task assigned to the UAV system (1) can be expressed by the following syntax:

[0041] Among them, the drone has sensors that can sense obstacles near it in real time online.

[0042] Define the drone’s surrounding perception radius as Circle All obstacles within can be perceived through sensor data. Assume that the obstacles perceived by the drone can be modeled as isolated non-intersecting circular regions. Define the set To complete φ s The set of obstacles perceived by the drone at time t during the process. The predicate function of the kth obstacle in is (2) With r o,s,k is φ s The center and radius of the kth obstacle in .

[0043] Define the drone to complete the target proposition φ goalThe obstacle proposition φ when safely avoiding all perceived obstacles during the process obs , that is, for any t∈[0,b K ], Assume that the initial state of the UAV system (1) x0: = x(0) does not belong to h o,s,k The union of the zero-level sets of (x(t)) in the target task φ s The kth target area in is π g,s,k When , the corresponding predicate function is formula (3) The target state is not time-varying, and the UAV system (1) exists to satisfy φ s The feasible solution of .

[0044] Therefore, the goal to be achieved by the system controller provided in this embodiment is: Established.

[0045] The following first describes the design of the obstacle avoidance controller in the system controller.

[0046] For each proposition φ s , and its artificial potential field function is defined as formula (5)U s (x) = U attr,s (x)+∑ k U rep,s,k (x); where U attr,s (x) is the gravitational potential field, For the drone in Proposition φ s The target state, π o,s,k is the set of obstacles perceived by the UAV at time t The predicate function h of the kth obstacle in o,s,k The subzero level set of (x(t)), With r o,s,k is φ s The center and radius of the kth obstacle in U rep,s,k (x) is φ s The repulsive potential field of the k-th obstacle in , The choice satisfies And the function (6) The zero-level set of k does not intersect, s ,η s,k is a positive scalar parameter.

[0047] Obstacle avoidance controller u cpf,s It is generated by solving the following quadratic programming problem: Equation (7) Among them, c sTo adjust the obstacle avoidance rate parameters, c s >0.

[0048] In order to avoid the UAV system (1) being trapped in the local minimum point, based on the solution of the above quadratic programming problem, the obstacle avoidance controller u cpf,s It is in the following form: in, To solve the above quadratic programming problem, we obtain the optimal controller; C g 、C d is the threshold of gradient and distance, C g >0,C d >0; δ is bounded perturbation, δ∈R m ; C g 、C d and δ satisfy the following conditions: ||g(x)δ||≤C δ , and the set In each time interval t∈[a s ,b s ] is not an invariant set.

[0049] Based on the above obstacle avoidance controller, let the target state be The maximum distance of the subzero level set is If the parameter k of the artificial potential field function s and η s,k satisfy: Then the obstacle avoidance controller u cpf,s The UAV system (1) can meet the safety requirements, namely

[0050] The following describes the design of the target arrival controller in the system controller.

[0051] For each proposition φ s , the target reaches the controller u cbf,s It is generated by solving the following quadratic programming problem: Equation (9) Where Q∈R m×m is a positive definite matrix, α s :(-∞,+∞)→(-∞,+∞) is the expansion Class function, satisfying α s (0)=0 and strictly monotonically increasing, is the time-varying robust control barrier function, Where η>0, is φ s The kth time-varying robust control barrier function in the present invention is described in the specification and claims using the subscript "s,k" to describe φ sFunction and parameter information of the kth time-varying robust control barrier function.

[0052] Among them, the time-varying robust control barrier function is in the following form: Among them, h g,s,k (x) is the predicate function corresponding to the target area; γ s,k (t) is a constraint parameter, which constrains h(x) according to the properties of the signal timing logic task, thereby completing the corresponding logic task.

[0053] The following describes the design of the system controller.

[0054] System Controller s Including the obstacle avoidance controller u cpf,s and the target arrival controller u cbf,s , which is of the following form: u s =λ s (x,h ob,s (x,t))u cbf,s +(1-λ s (x,h ob,s (x,t)))u cpf,s ; Among them, h ob,s (x) is the smooth lower approximation of the predicate function of all obstacles perceived by the drone, As shown in formula (6); s (x,h ob,s (x)) is defined as Ensure λ s (x,h ob,s (x,t))∈[0,1],β s is the decision function λ s (x,h ob,s (x,t)) decay rate constant.

[0055] The following describes the online parameter adjustment method of the above-mentioned drone system controller.

[0056] After obtaining the above system controller through the obstacle avoidance controller and the target reaching controller, since the system controller does not take into account the situation of facing unknown environment or limited control input conditions, the operation of the UAV is often difficult to ensure forward invariance, resulting in When the system returns to the collection The time within depends largely on the expansion Class function α s The selection of the parameters will affect the performance of the UAV mission.

[0057] In the target arrival controller, The initial state of the drone is x0 = x(t0), let Then γ s,k (t) Initialize and update the constraint parameters in step 3 based on the following rules: Formula (10)

[0058]

[0059]

[0060] In formula (10), ρ s,k is the robustness parameter, is the time parameter, Represents the predicate function h corresponding to the predicate logic μ g,s,k (x) In Proposition F [a,b] ψ, Represents the predicate function h corresponding to the predicate logic μ g,s,k (x) In Proposition G [a,b] In ψ, ψ is the area of ​​the signal timing logic task currently being performed by the UAV; Indicates When the predicate function h g,s,k The supremum of (x), r s,k is the robustness coefficient for the proposition, r s,k ∈R ＞0 , R ＞0 is a pre-set set of positive real numbers; Δ s,k To update the constraint parameters, Δ s,k ∈(0,γ s,k,∞ ). In order to avoid the Zeno phenomenon, the parameter I s,k are all positive real numbers, and l max >0 is Updated upper bound and in When belongs to γ s,k Parameters in (t).

[0061] In step 2, the time parameters are initialized as follows: when φ s =φ1, t0=0; when φ s When s≥2, t0=b s-1 .

[0062] In step 3, when the current time t is less than the time parameter and When t is less than or equal to 0, let t new =t, Update γ based on the above formula (10)s,k,0 , juxtapose h g,s,k (x0)=h g,s,k (x), at this time And according to Calculate γ s,k (t); among them, Greater than or equal to the preset l max hour,

[0063] In step 3, when the current time t is greater than or equal to the time parameter and When t is less than or equal to 0, let t new =t, update γ based on the update rule s,k,0 , juxtapose h g,s,k (x0)=h g,s,k (x), according to γ s,k (t) = (γ s,k,0 -γ s,k,∞ )exp(-l max (tt new ))+γ s,k,∞ Calculate γ s,k (t).

[0064] During the constraint parameter update process, if If it is greater than 0, the γ obtained after the last update is output s,k (t).

[0065] In step 4, if there is a moment In the time interval The constraint parameter γ s,k (t) No update occurs and in the time interval Internal s,k (t)>0, then output the constraint parameter γ obtained after the last update s,k (t).

[0066] The present invention provides an online parameter adjustment method for a drone system controller. This method generates an obstacle avoidance controller based on an existing artificial potential field function and a target reach controller based on an existing time-varying robust control obstacle function. These two functions are combined to form a system controller. The system controller controls a drone to complete a pre-set signal sequential logic task, enabling the drone to avoid any perceived obstacle when approaching it, and to fly toward the target location to complete the target sequential logic task when moving away from it. However, during actual research, the inventors discovered that relying solely on the aforementioned system controller, when faced with unknown environments or limited control input conditions, often fails to ensure forward invariance of the corresponding control obstacle function, thereby affecting the drone's task performance. To address this issue, the inventors implemented an online adjustment of the constraint parameters in the time-varying robust control obstacle function, enabling the drone to adjust the corresponding constraint parameters in real time based on its current state information and the nature of the signal sequential logic task. This allows the drone to better complete complex signal sequential logic tasks under unknown environmental disturbances or limited control input conditions.

[0067] To further illustrate the effectiveness of the above-mentioned method for online parameter adjustment of the drone system controller, the present invention conducts the following experiments and analyzes the experimental results through Figures 2 to 6 Provide explanation.

[0068] The drone used in this experiment is equipped with multiple optical motion capture light balls and a 360° two-dimensional laser radar. The optical motion capture light balls are used to calculate and obtain global information about the drone within the VICON optical motion capture system, while the two-dimensional laser radar is used to detect nearby obstacles. The axial (forward, backward, left, and right) velocity command serves as the drone's control input, u.

[0069] Experimental scenario such as Figure 2 As shown, Figure 2 The global coordinates XYZ are shown in the figure. There are two obstacles on the experimental platform, and their global positions are [0.12, 3.61] T ,[0.27,-0.56] T , corresponding to radii of 0.36m and 0.56m respectively. The drone has no prior knowledge of obstacles at the initial moment and needs to perceive surrounding obstacle information in real time and ensure the safety of the drone itself when completing the signal timing logic task. The signal timing logic task of this experiment is assigned to the drone after the drone rises and stabilizes to the desired height. The time when the drone starts to execute the signal timing logic task is defined as t0. The target arrival signal timing logic task assigned to the drone in this experiment is as follows: φ1 = G [15,20] (||x-[2,3.22] T||≤0.5);φ2=F [40,60] (||x-[-2,3.17] T ||≤0.5);φ3=F [110,120] (||x-[1,-3.17] T ||≤0.5).

[0070] The trajectory of the drone in this experiment under the VICON system is as follows Figure 3 As shown, the red circle is the obstacle, and the blue circle is the target arrival signal timing logic task assigned to the UAV. It can be seen that the security of the system is met, that is, The corresponding h(x)(h g,1,1 ,h g,2,1 ,h g,3,1 ),γ(t)(γ 1,1 ,γ 2,1 ,γ 3,1 ) and the control barrier function Respectively as Figure 4 、 5 As shown. It can be seen that when the UAV is completing the obstacle avoidance task, even if u cbf,s in u s The UAV system can still ensure the forward invariance of the system control obstacle function through online adjustment of parameters. Based on the image of γ(t), it can be seen that the target mission of the UAV has been completed, that is, The speed control command u of the drone is as follows Figure 6 shown.

[0071] Through the above experiments, the effectiveness of the online parameter adjustment method of the drone system controller provided by the present invention is further verified. By adjusting the relevant parameters in the system controller online, the drone can better complete signal timing logic tasks with complex requirements under the disturbance of unknown environment or limited control input conditions.

[0072] The above description and embodiments are intended to explain the scope of protection of the present invention, but do not constitute a limitation thereto. Modifications, equivalent substitutions, or other improvements to the embodiments of the present invention or portions thereof that can be obtained by a person of ordinary skill in the art through logical analysis, reasoning, or limited experimentation based on the teachings of the present invention or the above embodiments, combined with common knowledge, ordinary technical knowledge in the field, and / or prior art, should all be included within the scope of protection of the present invention.

Claims

1. A method for online parameter adjustment of an unmanned aerial vehicle (UAV) system controller, wherein the UAV performs a signal timing logic task under the control of the system controller; the system controller includes an obstacle avoidance controller generated based on an artificial potential field function and a target arrival controller generated based on a time-varying robust control obstacle function; wherein the method comprises: In the target arrival controller, the time-varying robust control obstacle function includes a predicate function corresponding to the target area and constraint parameters determined based on the signal sequential logic task; During the operation of the UAV, the constraint parameters are updated in real time through the following steps: Step 1: Obtain the predicate function and robustness parameters corresponding to the target area; Step 2: Initialize time parameters and constraint parameters based on the nature of the signal timing logic task currently being performed by the UAV; Step 3: When the time-varying robust control barrier function is non-positive, updating the constraint parameters according to a preset update rule; Step 4: Within a preset period, determine whether the time-varying robust control barrier function corresponding to the constraint parameter is less than or equal to 0. If so, return to step 3 to continue updating the constraint parameter; otherwise, maintain the constraint parameter obtained after the last update; Among them, the time-varying robust control barrier function is given by In other words, in the proposition φ s The kth time-varying robust control barrier function is The predicate function corresponding to the target area is h g,s,k (x), the constraint parameter is γ s,k (t), and among them,c s,k (t)=(γ s,k,0 -c s,k,∞ )exp(-I s,k (t-t0))+γ s,k,∞ ; The initial state of the drone is x0=x(t0), let Then γ s,k (t) Initialize and update the constraint parameters in step 3 based on the following rules: In the above rules, ρ s,k is the robustness parameter, ρ s,k >0; t0 is each proposition φ s The initial moment, is the time parameter, Represents the predicate function h corresponding to the predicate logic μ g,s,k (x) In Proposition F [a,b] ψ, Represents the predicate function h corresponding to the predicate logic μ g,s,k (x) in proposition g [a,b] In ψ, ψ is the area of ​​the signal timing logic task currently being performed by the UAV; Indicates When the predicate function h g,s,k The supremum of (x), r s,k is the robustness coefficient for the proposition, r s,k ∈R ＞0 , R ＞0 is a pre-set set of positive real numbers; Δ s,k To update the constraint parameters, Δ s,k ∈(0,γ s,k,∞ ).

2. The method for online parameter adjustment of a drone system controller according to claim 1, wherein In the second step, the time parameters are initialized as follows: s =φ1, t0=0; when φ s When s≥2, t0=b s-1 .

3. The method for online parameter adjustment of a drone system controller according to claim 2, wherein In step 3, when the current time t is less than the time parameter and When t is less than or equal to 0, let t new =t, Update γ based on the update rule s,k,0 , juxtapose h g,s,k (x0)=h g,s,k (x), at this time And according to γ s,k (t) = (γ s,k,0 -γ s,k,∞ )exp(-I s,k (tt new ))+γ s,k,∞ Calculate γ s,k (t); where s,k Greater than or equal to the preset l max When s,k =l max .

4. The method for online parameter adjustment of a drone system controller according to claim 3, wherein In step 3, the current time t is greater than or equal to the time parameter And b s When (x,t) is less than or equal to 0, let t new =t, update γ based on the update rule s,k,0 , juxtapose h g,s,k (x0)=h g,s,k (x), according to γ s,k (t) = (γ s,k,0 -γ s,s,∞ )exp(-l max (tt new ))+γ s,k,∞ Calculate γ s,k (t).

5. The method for online parameter adjustment of a drone system controller according to claim 4, wherein: During the constraint parameter updating process, if If it is greater than 0, the γ obtained after the last update is kept s,k (t).

6. The method for online parameter adjustment of a drone system controller according to claim 1, wherein: The obstacle avoidance controller is generated based on an artificial potential field function; For each proposition φ s , its artificial potential field function is defined as U s (x) = U attr,s (x)+∑ k U rep,s,k (x); where U attr,s (x) is the gravitational potential field, For the drone in Proposition φ s The target state, π o,s,k is the set of obstacles perceived by the UAV at time t The predicate function h of the kth obstacle in o,s,k The subzero level set of (x(t)), With r o,s,k is φ s The center and radius of the kth obstacle in U rep,s,k (x) is φ s The repulsive potential field of the k-th obstacle in , The choice satisfies And function The zero-level set of k does not intersect, s ,η s,k is a positive scalar parameter; The obstacle avoidance controller u cpf,s Generated by solving the following quadratic programming problem: Among them, c s To adjust the obstacle avoidance rate parameters, c s >0; Based on the solution of the above quadratic programming problem, the obstacle avoidance controller u cpf,s In the following form: in, To solve the above quadratic programming problem, we obtain the optimal controller; C g 、C d is the threshold of gradient and distance, C g >0,C d >0; δ is bounded perturbation, δ∈R m ; C g 、C d and δ satisfy the following conditions: ||g(x)δ||≤C δ , and the set In each time interval t∈[a s ,b s ] is not an invariant set.

7. The method for online parameter adjustment of a drone system controller according to claim 6, wherein: The target arrival controller is generated based on a time-varying robust control obstacle function; For each proposition φ s , the target reaches the controller u cbf,s Generated by solving the following quadratic programming problem: Where Q∈R m×m is a positive definite matrix, α s :(-∞,+∞)→(-∞,+∞) is the expansion Class function, satisfying α s (0)=0 and strictly monotonically increasing, is the time-varying robust control barrier function, Where η>

0.

8. The method for online parameter adjustment of a UAV system controller according to claim 7, wherein: The system controller u s Including the obstacle avoidance controller u cpf,s and the target arrival controller u cbf,s , which is of the following form: u s =λ s (x,h ob,s (x,t))u cbf,s +(1-λ s (x,h ob,s (x,t)))u cpf,s ; Among them, h ob,s (x) is the smooth lower approximation of the predicate function of all obstacles perceived by the drone, η ob >0,λ s (x,h ob,s (x)) is defined as Ensure λ s (x,h ob,s (x,t))∈[0,1],β s is the decision function λ s (x,h ob,s (x,t)) decay rate constant.

9. The method for online parameter adjustment of a drone system controller according to claim 8, wherein: In step 4, for any k, if there is a time In the time interval The constraint parameter γ s,k (t) No update occurs and in the time interval Internal s,k (t)>0, determine each sequential logic task φ s All completed.

Images