Method for determining control instruction of wingman in unmanned aerial vehicle formation and terminal device

Abstract

The application is suitable for the technical field of unmanned aerial vehicle control, and provides a method for determining control instructions of wingmen in an unmanned aerial vehicle formation and a terminal device, wherein the method obtains intermediate state variables of a lead aircraft and wingmen under the action of control variables according to a motion model when the unmanned aerial vehicle formation flies, and initial state variables of the lead aircraft and the wingmen which are predefined; calculates state errors of the wingmen relative to the lead aircraft according to the intermediate state variables of the lead aircraft and the wingmen; divides language variables of a predefined state error domain and a control instruction error domain to obtain a state error language variable set and a control instruction error language variable set; determines control instruction fuzzy constraints according to membership functions of language variable values in each language variable set and fuzzy control rules; and determines control instructions of the wingmen based on a DMPC method and the control instruction fuzzy constraints. The application can reduce the consumption of wingman control energy.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of unmanned aerial vehicle control, and particularly relates to a method for determining control instructions of a subordinate unmanned aerial vehicle in an unmanned aerial vehicle formation and a terminal device. BACKGROUND

[0002] In recent years, cluster cooperation of fixed-wing unmanned aerial vehicles (UAVs) has become an important development trend of unmanned aerial vehicle system applications. Small multi-unmanned aerial vehicles have attracted much attention due to their high cooperation, multi-task and low cost. Unmanned aerial vehicle formation control technology, as a sub-problem of multi-unmanned aerial vehicle cooperation, has been widely studied. The goal of unmanned aerial vehicle formation flight is to achieve an ideal formation by controlling the behavior of each unmanned aerial vehicle.

[0003] Common formation flight strategies include master-slave method, virtual structure method, behavior-based method and consensus theory. Unmanned aerial vehicle formations can be divided into centralized and distributed types according to communication methods. In the centralized method, a control center is established in the formation to control the entire formation, while in the distributed method, the overall formation is completed through information exchange between adjacent individuals. Distributed control does not have a control center, and thus has better flexibility and scalability than centralized control. For formation control methods, common methods include consensus control, adaptive feedback, sliding mode control and fuzzy control. However, these methods are not convenient for handling constraints, and optimization methods such as distributed model predictive control (DMPC) can be used to design cost functions and constraint conditions according to specific requirements.

[0004] For unmanned aerial vehicle formation control problems, the stability and feasibility of the formation are generally studied, and little research is done on energy saving during the formation process. Most existing technologies set parameters in the DMPC algorithm, increase the weight of the control quantity to reduce the effect of the control quantity, but this slows down the state convergence rate and increases energy consumption. SUMMARY

[0005] Embodiments of the present application provide a method for determining control instructions of a subordinate unmanned aerial vehicle in an unmanned aerial vehicle formation and a terminal device, which can solve the problem of large energy consumption of the subordinate unmanned aerial vehicle in the formation.

[0006] In a first aspect, embodiments of the present application provide a method for determining control instructions of a subordinate unmanned aerial vehicle in an unmanned aerial vehicle formation, comprising:

[0007] According to the motion model of the unmanned aerial vehicle formation flight, the pre-defined initial state quantity of the leader and the pre-defined initial state quantity of the subordinate, the intermediate state quantity of the leader and the intermediate state quantity of the subordinate are obtained under the action of the control quantity.

[0008] According to the intermediate state quantity of the leader and the intermediate state quantity of the deputy, a state error of the deputy relative to the leader is calculated;

[0009] Language variable partition is performed on a predefined state error universe of the deputy and a predefined control instruction error universe to obtain a set of state error language variables of the deputy and a set of control instruction error language variables of the deputy;

[0010] According to a membership function corresponding to each language variable value in the set of state error language variables, a membership function corresponding to each language variable value in the set of control instruction error language variables, and a predefined control instruction fuzzy control rule, a control instruction fuzzy constraint of the deputy is determined;

[0011] Based on a distributed model predictive control method, a control instruction of the deputy is determined according to the control instruction fuzzy constraint of the deputy.

[0012] Optionally, an expression of a motion model of the unmanned aerial vehicle formation in flight is as follows:

[0013]

[0014]

[0015]

[0016] wherein (x i , y i ) represents coordinates of the i th deputy in the inertial coordinate system in the unmanned aerial vehicle formation, ψ i represents a heading angle of the i th deputy in the unmanned aerial vehicle formation, V i represents a speed of the i th deputy in the unmanned aerial vehicle formation, ω i represents an angular speed of the i th deputy in the unmanned aerial vehicle formation, ψ i c represents a heading angle control instruction of the i th deputy in the unmanned aerial vehicle formation, V i c represents a speed control instruction of the i th deputy in the unmanned aerial vehicle formation, and both represent inertial time constants of the autopilot heading angle channel, τ v represents an inertial time constant of the speed channel, i = 1, 2, …, N a , N a represents a total number of the deputies in the flight formation, V min represents a minimum flight speed of the deputy, V max represents a maximum flight speed of the deputy, a max represents a maximum angular speed of the deputy, ω max represents a maximum angular speed of the deputy, α max represents a maximum angular acceleration of the deputy, represents the angular acceleration of the i th wingman, represents the acceleration of the i th wingman.

[0017] Optionally, the intermediate state quantity X0 of the leader is [x, y, ψ, V, ω] T .

[0018] Optionally, the intermediate state quantity X of the wingman is i = [x i , y i , ψ i , V i , ω i ] T ; wherein X i represents the intermediate state quantity of the i th wingman.

[0019] Optionally, according to the intermediate state quantity of the leader and the intermediate state quantity of the wingman, the state error of the wingman relative to the leader is calculated, including:

[0020] Taking the north-east coordinate system P n OP e as the inertial coordinate system and the leader as the reference point R, the state quantity X R of the reference point R in the inertial coordinate system is obtained, X R = [x R , y R , ψ R , V R , ω R ] T ; the position of the reference point and the state quantity of the reference point are determined by the state quantity of the leader;

[0021] A reference point coordinate system X R OY R is established in the inertial coordinate system with the leader as the reference point, and the coordinates of the target point G i of the i th wingman in the reference point coordinate system X R OY R are determined

[0022] The state quantity of the target point G i of the i th wingman in the inertial coordinate system is obtained through the calculation formula

[0023]

[0024] wherein,

[0025] The state quantity of the target point G i of the i th wingman in the inertial coordinate system is obtained through the calculation formula

[0026]

[0027] state error represents the state error of the ith wingman relative to the lead aircraft in the reference point coordinate system; wherein, represents the coordinate transformation matrix from the inertial frame to the reference frame, x i , y i , ψ i , V i , and ω i all represent the state quantities of the ith wingman in the reference point coordinate system, represents the error of the ith wingman relative to the lead aircraft in the x direction in the reference point coordinate system, represents the error of the ith wingman relative to the lead aircraft in the y direction in the reference point coordinate system, represents the heading angle error of the ith wingman relative to the lead aircraft in the reference point coordinate system, represents the speed error of the ith wingman relative to the lead aircraft in the reference point coordinate system, represents the angular velocity error of the ith wingman relative to the lead aircraft in the reference point coordinate system.

[0028] Optionally, the control instruction error includes a speed control instruction error of the ith wingman and a heading angle control instruction error of the ith wingman

[0029] Optionally, the linguistic variables include NB, NM, MS, Z, PS, PM, and PB; wherein, NB represents a large negative error, NM represents a medium negative error, MS represents a small negative error, Z represents almost no error, PS represents a small positive error, PM represents a medium positive error, and PB represents a large positive error.

[0030] Optionally, the state error linguistic variable set represents the state error linguistic variable set of the ith wingman; wherein,

[0031] Optionally, the control instruction error linguistic variable set represents the control instruction error linguistic variable set of the ith wingman; wherein,

[0032] Optionally, according to the membership function corresponding to each linguistic variable value in the state error linguistic variable set, the membership function corresponding to each linguistic variable value in the control instruction error linguistic variable set, and the pre-defined control instruction fuzzy control rule, the control instruction fuzzy constraint of the wingman is determined, including:

[0033] According to the membership function corresponding to each language variable value in the state error language variable set, a state error membership image is drawn, and a first language variable membership set corresponding to the state error is obtained; the first language variable membership set includes the membership corresponding to each language variable value in the state error language variable set;

[0034] According to the membership function corresponding to each language variable value in the control instruction error language variable set, a control instruction error membership image is drawn, and a second language variable membership value corresponding to the control instruction error is obtained; the second language variable membership set includes the membership corresponding to each language variable value in the control instruction error language variable set;

[0035] According to the first language variable membership set, the second language variable membership set, and the pre-defined fuzzy control rule, a plurality of control instruction error fuzzy sets are obtained;

[0036] The plurality of control instruction error fuzzy sets are aggregated to obtain a control instruction error aggregated fuzzy set;

[0037] From the control instruction error aggregated fuzzy set, an element set corresponding to the maximum membership is selected as a maximum control instruction error fuzzy set, and a language variable value corresponding to the maximum control instruction error fuzzy set is taken as an optimal language variable;

[0038] According to the optimal language variable, a boundary of the wingman control instruction fuzzy constraint is determined, and a control instruction fuzzy constraint of the wingman is obtained.

[0039] Optionally, the control instruction fuzzy control rule includes:

[0040] According to and the speed control instruction error of the i th wingman is determined

[0041] According to and the heading angle control instruction error of the i th wingman is determined

[0042] Optionally, according to the optimal language variable, a boundary of the wingman control instruction fuzzy constraint is determined, and a control instruction fuzzy constraint of the wingman is obtained, including:

[0043] For the i th wingman, the control instruction corresponding to the same membership as the optimal language variable is taken as the left boundary of the wingman control instruction fuzzy constraint

[0044] For the i-th wingman, the control command corresponding to the position where the value of the linguistic variable to the right of the optimal linguistic variable is the same as the membership degree of the optimal linguistic variable is taken as the right boundary of the fuzzy constraint of the wingman's control command.

[0045] According to the left boundary and right boundary Determine the fuzzy constraints of the control commands for the i-th wingman. Among them, U X Indicates the instruction space.

[0046] Optionally, based on a distributed model predictive control method, the control commands of the wingman are determined according to the fuzzy constraints of the wingman's control commands, including:

[0047] Through calculation formula

[0048]

[0049] Obtain the coordinates of the i-th wingman relative to the j-th wingman in the reference point coordinate system X. R OY R State error in Where j = 1, 2, ..., N a And j≠i;

[0050] Through calculation formula

[0051]

[0052]

[0053] Xe ij (R) (τ|t k )=Xe j (τ|t k )-Xe i (τ|t k )

[0054]

[0055]

[0056] Obtain the control command trajectory of the i-th wingman. And the wingman in t k Control commands at +δ time This is determined as the final control command; where δ represents the update cycle, and t... k This represents the time corresponding to the actual state quantity after k updates. Let τ represent the set of natural numbers, τ∈{t} k+ δs, s = 0, 1, K, N p -1, N p denotes the prediction step, δN p denotes the prediction horizon, S (X) and S (U) both denote the normalization matrix, M1, M2, N all denote positive definite symmetric matrices, X i denotes the state of the i-th wingman, X -i denotes the state of the wingman adjacent to the i-th wingman, X i (t k ) denotes the state of the i-th wingman at time t k , U i (t k ) denotes the control of the i-th wingman at time t k , denotes the predicted state trajectory of the i-th wingman, denotes the estimated state trajectory of the i-th wingman within the prediction horizon, denotes the optimal state trajectory of the i-th wingman within the prediction horizon, denotes the predicted control trajectory of the i-th wingman, denotes the estimated control trajectory of the i-th wingman, denotes the optimal predicted control trajectory of the i-th wingman, denotes the state transition equation of the i-th wingman from time τ to time τ + δ, denotes the predicted state of the i-th wingman at time τ + δ, denotes the estimated state of the i-th wingman at time τ + δ, denotes the optimal predicted state of the i-th wingman at time τ + δ, denotes the optimal predicted state of the adjacent wingman of the i-th wingman at time τ + δ. denotes the state error of wingman i relative to the lead aircraft at time τ, A(τ|t k ) denotes the coordinate transformation matrix from the inertial frame to the reference frame at time τ, X i (τ|t k ) denotes the state of the i-th wingman at time τ, denotes the state of the target position of the i-th wingman at time τ, denotes the control of the target position of the i-th wingman at time τ, denotes the state error of the j-th wingman relative to the lead aircraft at time τ, denotes the estimated state of the adjacent wingman of the i-th wingman at time t k , denotes the estimated control of the adjacent wingman of the i-th wingman at time t k , represents the state error of the ith follower and the jth follower at the time τ, represents the rank of the transformation of represents the control error of the follower i relative to the leader at the time τ, ||·|| represents a norm.

[0057] In a second aspect, an embodiment of the present application provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method for determining the control instruction of the follower in the UAV formation when executing the computer program.

[0058] The above scheme of the present application has the following beneficial effects:

[0059] In some embodiments of the present application, according to the motion model of the UAV formation flight, the pre-defined initial state quantity of the leader and the pre-defined initial state quantity of the follower, the intermediate state quantity of the leader and the intermediate state quantity of the follower are obtained under the action of the control quantity, then the state error of the follower relative to the leader is calculated according to the intermediate state quantity of the leader and the intermediate state quantity of the follower, the pre-defined state error domain of the follower and the pre-defined control instruction error domain are divided into language variables, the state error language variable set of the follower and the control instruction error language variable set of the follower are obtained, then the control instruction fuzzy constraint of the follower is determined according to the membership function corresponding to each language variable value in the state error language variable set, the membership function corresponding to each language variable value in the control instruction error language variable set and the pre-defined control instruction fuzzy control rule, and finally the control instruction of the follower is determined based on the distributed model predictive control method according to the control instruction fuzzy constraint of the follower. Wherein, by determining the control instruction fuzzy constraint of the follower, and then determining the control instruction of the follower based on the distributed model predictive control method, the amplitude change of the control quantity can be reduced under the condition that the control quantity weight is unchanged, thereby reducing the energy consumption of the follower control.

[0060] Other beneficial effects of the present application will be described in detail in the subsequent specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creating any creative labor.

[0062] Figure 1 The flow chart of the method for determining the control instruction of the follower in the UAV formation provided by an embodiment of the present application; ​

[0063] Figure 2 A schematic diagram of formation flight of UAVs is provided for an embodiment of the present application;

[0064] Figure 3a A linguistic variable membership image of y-direction error of wingman is provided for an embodiment of the present application;

[0065] Figure 3b A linguistic variable membership image of x-direction error of wingman is provided for an embodiment of the present application;

[0066] Figure 3c A linguistic variable membership image of heading angle error of wingman is provided for an embodiment of the present application;

[0067] Figure 3d A linguistic variable membership image of speed error of wingman is provided for an embodiment of the present application;

[0068] Figure 4a A linguistic variable membership image of angle command error of wingman is provided for an embodiment of the present application;

[0069] Figure 4b A linguistic variable membership image of speed command error of wingman is provided for an embodiment of the present application;

[0070] Figure 5a A schematic diagram of flight trajectory of wingman when DMPC method is adopted is provided for an embodiment of the present application;

[0071] Figure 5b A schematic diagram of flight trajectory of wingman when fuzzy constrained DMPC method is adopted is provided for an embodiment of the present application;

[0072] Figure 6a A comparison schematic diagram of x-direction distance variation of DMPC method and fuzzy constrained DMPC method is provided for an embodiment of the present application;

[0073] Figure 6b A comparison schematic diagram of y-direction distance variation of DMPC method and fuzzy constrained DMPC method is provided for an embodiment of the present application;

[0074] Figure 6c A comparison schematic diagram of speed variation of DMPC method and fuzzy constrained DMPC method is provided for an embodiment of the present application;

[0075] Figure 6d A comparison schematic diagram of heading angle variation of DMPC method and fuzzy constrained DMPC method is provided for an embodiment of the present application;

[0076] Figure 7 A comparison schematic diagram of energy loss of DMPC method and fuzzy constrained DMPC method is provided for an embodiment of the present application;

[0077] Figure 8a A DMPC method provided by an embodiment of the present application and a flight distance difference diagram of a fuzzy constraint DMPC method;

[0078] Figure 8b A DMPC method provided by an embodiment of the present application and a speed change cumulative difference diagram of a fuzzy constraint DMPC method;

[0079] Figure 8c A DMPC method provided by an embodiment of the present application and an angular velocity change cumulative difference diagram of a fuzzy constraint DMPC method;

[0080] Figure 8d A DMPC method provided by an embodiment of the present application and an energy loss difference diagram of a fuzzy constraint DMPC method;

[0081] Figure 9 A terminal device structure diagram provided by an embodiment of the present application. DETAILED DESCRIPTION

[0082] In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular sequences of steps, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known methods, devices, circuits, and

[0083] It will be understood that the term "includes," "including," "has," "having," "comprises," "comprising" or "contains," "containing" when used in this specification and in the following claims, specifies the presence of the stated features, integers, steps, operations, elements, and / or components but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0084] It will be understood that the term "and / or," when used in this specification and in the following claims, can encompass the meaning of "and" and / or the meaning of "or." Similarly, the term "and / or" when used in the context of the following claims can encompass the meaning of "and" and / or the meaning of "or."

[0085] As used in this specification and claims, the terms "if" and "when" can be interpreted to mean "upon" or "in response to determining," or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [a described condition or event] is detected" can be interpreted to mean "upon determining" or "in response to determining," or "upon detecting [the described condition or event]" or "in response to detecting [the described condition or event]," depending on the context.

[0086] In addition, in the description of the present application and the appended claims, the terms "first", "second", "third", etc. are used only to distinguish descriptions and cannot be understood as indicating or implying relative importance.

[0087] Reference to "one embodiment" or "some embodiments" or "one implementation" or "some implementations" or "one example" or "some examples" described in the present application means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment or implementation of the present application. Therefore, the phrases "in one embodiment", "in some embodiments", "in other some embodiments", "in yet some embodiments", etc. appearing in various places in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically stated. The terms "comprise", "include", "have" and their variants mean "including but not limited to", unless otherwise specifically stated.

[0088] In view of the problem of large energy consumption of wingman control in current UAV formation, the present application provides a method for determining wingman control instructions in UAV formation and a terminal device, wherein the method obtains intermediate state quantities of the lead aircraft and intermediate state quantities of the wingman according to a motion model of the UAV formation in flight, pre-defined initial state quantities of the lead aircraft and pre-defined initial state quantities of the wingman under the action of control quantities, then calculates state error of the wingman relative to the lead aircraft according to the intermediate state quantities of the lead aircraft and the intermediate state quantities of the wingman, divides the pre-defined state error universe of the wingman and the pre-defined control instruction error universe by language variables to obtain a set of state error language variables of the wingman and a set of control instruction error language variables of the wingman, then determines the control instruction fuzzy constraint of the wingman according to the membership function corresponding to each language variable value in the set of state error language variables, the membership function corresponding to each language variable value in the set of control instruction error language variables and the pre-defined control instruction fuzzy control rule, and finally determines the control instruction of the wingman based on the distributed model predictive control method according to the control instruction fuzzy constraint of the wingman. Wherein, by determining the control instruction fuzzy constraint of the wingman and then determining the control instruction of the wingman based on the distributed model predictive control method, the amplitude change of the control quantity can be reduced under the condition that the weight of the control quantity remains unchanged, thereby reducing the energy consumption of wingman control.

[0089] As shown in Figure 1 The method for determining wingman control instructions in UAV formation provided by the present application mainly includes the following steps:

[0090] Step 11, obtaining intermediate state quantities of the lead aircraft and intermediate state quantities of the wingman according to a motion model of the UAV formation in flight, pre-defined initial state quantities of the lead aircraft and pre-defined initial state quantities of the wingman under the action of control quantities.

[0091] The UAV formation described above is composed of a leader and wingmen. In an embodiment of the present application, the UAV formation includes one leader and multiple wingmen. In another embodiment of the present application, the UAV formation includes one leader and one wingman. The wingman refers to a UAV that follows the leader to perform a task in formation flight, and the leader refers to a UAV that leads the formation in formation flight.

[0092] In some embodiments of the present application, the initial state of the leader and the initial state of the wingmen can be given initial values according to human experience when determining the flight formation motion model.

[0093] In an embodiment of the present application, a two-dimensional planar kinematic model of a fixed-wing UAV is used to describe the motion model of the UAV formation in flight.

[0094] Specifically, the expression of the motion model of the UAV formation in flight is as follows:

[0095]

[0096] where (x i , y i ) represents the coordinates of the i-th wingman in the inertial coordinate system, ψ i represents the heading angle of the i-th wingman, V i represents the speed of the i-th wingman in the UAV formation, and ω i represents the angular velocity of the i-th wingman in the UAV formation.

[0097] For the control of the wingmen, in an embodiment of the present application, an autopilot system is used to control the position and attitude of the wingmen. The expression of the autopilot model is as follows:

[0098]

[0099] where ψ i c represents the heading angle control command of the i-th wingman, V i c represents the speed control command of the i-th wingman, and both represent the inertial time constant of the heading angle channel of the autopilot, τ v represents the inertial time constant of the speed channel, i = 1, 2, …, N a , and N a represents the total number of wingmen in the flight formation.

[0100] Considering the state constraints in the actual flight of the UAV, including speed constraints, acceleration constraints, angular velocity constraints, and angular acceleration constraints, the state constraint expression of the i-th wingman is as follows:

[0101]

[0102] Among them, V min V represents the minimum flight speed of the wingman. max Indicates the maximum flight speed of the wingman, a max ω represents the maximum acceleration of the wingman. max α represents the maximum angular velocity of the wingman. max This indicates the wingman's maximum angular acceleration. Let represent the angular acceleration of the i-th wingman. Indicates the first i The acceleration of the wingman.

[0103] Based on the above motion model, the intermediate state variables of the lead aircraft can be determined as X0 = [x, y, ψ, V, ω]. T The intermediate state quantity X of the wingman i =[x i y i , ψ i V i ω i ] T Among them, X i This represents the intermediate state quantity of the i-th wingman.

[0104] Step 12: Calculate the state error of the wingman relative to the lead aircraft based on the intermediate state quantities of the lead aircraft and the wingman.

[0105] Calculating the state error of the wingman relative to the lead aircraft in this step is to prepare for determining the fuzzy constraints of the wingman's control commands.

[0106] Step 13: Perform linguistic variable partitioning on the predefined wingman's state error domain and the predefined control command error domain to obtain the wingman's state error linguistic variable set and the wingman's control command error linguistic variable set.

[0107] It should be noted that the aforementioned wingman's state error domain and control command error domain represent the wingman's state error range and control command error range, respectively.

[0108] In the embodiments of this application, the aforementioned control command error includes the speed control command error of the i-th wingman. Error of heading angle control command with the i-th wingman

[0109] The processing procedure of the language variable division is the process of fuzzifying the state error universe and the control instruction error universe of each wingman, and the language variables include NB, NM, MS, Z, PS, PM and PB. Among them, NB represents a large negative error, NM represents a medium negative error, MS represents a small negative error, Z represents almost no error, PS represents a small positive error, PM represents a medium positive error, and PB represents a large positive error.

[0110] Specifically, after the language variable division, the state error language variable set of the ith wingman is obtained denotes the state error language variable set of the ith wingman; wherein,

[0111] Similarly, the control instruction error language variable set of the ith wingman is denotes the control instruction error language variable set of the ith wingman; wherein,

[0112] Step 14, according to the membership function corresponding to each language variable value in the state error language variable set, the membership function corresponding to each language variable value in the control instruction error language variable set, and the pre-defined control instruction fuzzy control rule, the control instruction fuzzy constraint of the wingman is determined.

[0113] It should be noted that when the language variable set is determined, the membership function corresponding to each language variable value is also determined. Generally, the membership function includes a Z-type membership function, a Gaussian-type membership function and an S-type membership function, and each of the above membership functions belongs to the public knowledge and will not be described here.

[0114] Step 15, based on the distributed model predictive control method, the control instruction of the wingman is determined according to the control instruction fuzzy constraint of the wingman.

[0115] The distributed model predictive control method is an effective method for solving large-scale system control problems, and its advantages are: (1) reducing the calculation burden of each subsystem; (2) improving the scalability of the system under multiple controllers; (3) strong fault tolerance of the system, etc.

[0116] In the embodiment of the present application, by adding the control instruction fuzzy constraint to the DMPC method, a fuzzy constraint DMPC controller is constructed to determine the control instruction of the wingman, which can reduce the amplitude change of the control quantity and make the control quantity in a constant state within a period of time under the condition that the weight of the wingman control quantity is unchanged, thereby achieving the effect of reducing the wingman control energy consumption.

[0117] The following is an illustrative example of the specific process of step 12 (calculating the state error of the wingman relative to the lead aircraft based on the intermediate state quantities of the lead aircraft and the wingman).

[0118] Step 12.1, set the northeast coordinate system P n OP e Using the inertial coordinate system as the reference point R, we obtain the state quantity X of the reference point R in the inertial coordinate system. R X R =[x R y R , ψ R V R ω R ] T .

[0119] The location of the aforementioned reference point and the state variables of the reference point are determined by the state variables of the lead machine.

[0120] Step 12.2: Establish a reference point coordinate system X in the inertial coordinate system with the lead aircraft as the reference point. R OY R And determine the target point G of the i-th wingman. i In the reference point coordinate system X R OY R coordinates in

[0121] The flight diagram of the i-th wingman after establishing the inertial coordinate system and the reference point coordinate system is as follows: Figure 2 As shown, Figure 2 The horizontal axis represents the coordinates in the east direction of the inertial coordinate system, and the vertical axis represents the coordinates in the north direction of the inertial coordinate system. G i G1, G2, G3, and G5 all represent target points, UAV i This represents the i-th wingman.

[0122] Step 12.3, using the calculation formula

[0123]

[0124] Obtain the target point G of the i-th wingman in the inertial coordinate system. i state variables

[0125] in, ψx=ψ R V G =V R ω G =ω R .

[0126] Step 12.4, using the calculation formula

[0127]

[0128] state error

[0129] In the above formula, represents the state error of the ith wingman relative to the lead aircraft in the reference point coordinate system; wherein, represents the coordinate transformation matrix from the inertial system to the reference system, x i , y i , ψ i , V i and ω i all represent the state quantities of the ith wingman in the reference point coordinate system, represents the error of the ith wingman relative to the lead aircraft in the x direction in the reference point coordinate system, represents the error of the ith wingman relative to the lead aircraft in the y direction in the reference point coordinate system, represents the heading angle error of the ith wingman relative to the lead aircraft in the reference point coordinate system, represents the speed error of the ith wingman relative to the lead aircraft in the reference point coordinate system, represents the angular velocity error of the ith wingman relative to the lead aircraft in the reference point coordinate system.

[0130] The specific process of step 14 (determining the control command fuzzy constraint of the wingman according to the membership function corresponding to each language variable value in the state error language variable set, the membership function corresponding to each language variable value in the control command error language variable set, and the pre-defined control command fuzzy control rule) is described exemplarily as follows.

[0131] Step 14.1, according to the membership function corresponding to each language variable value in the state error language variable set, a state error membership graph is drawn to obtain a first language variable membership set corresponding to the state error.

[0132] The above first language variable membership set includes the membership corresponding to each language variable value in the state error language variable set.

[0133] Figure 3a is the membership graph of the y direction error, the horizontal coordinate of which represents the state error of the ith wingman in the y direction, and the vertical coordinate represents the membership.

[0134] Figure 3b is the membership graph of the x direction error, the horizontal coordinate of which represents the state error of the ith wingman in the x direction, and the vertical coordinate represents the membership.

[0135] Figure 3cThe membership image of the heading angle error, whose abscissa represents the heading angle error of the ith wingman, and whose ordinate represents the membership.

[0136] Figure 3d The membership image of the speed error, whose abscissa represents the speed error of the ith wingman, and whose ordinate represents the membership.

[0137] For example, the first linguistic variable membership set of the ith wingman can be represented as

[0138]

[0139] Step 14.2, according to the membership function corresponding to each linguistic variable value in the control instruction error linguistic variable set, a control instruction error membership image is drawn, and a second linguistic variable membership value corresponding to the control instruction error is obtained.

[0140] The above second linguistic variable membership set includes the membership corresponding to each linguistic variable value in the control instruction error linguistic variable set.

[0141] Figure 4a The membership image of the heading angle error, whose abscissa represents the heading angle error of the ith wingman, and whose ordinate represents the membership.

[0142] Figure 4b The membership image of the speed error, whose abscissa represents the speed error of the ith wingman, and whose ordinate represents the membership.

[0143] For example, the second linguistic variable membership set of the ith wingman can be represented as

[0144]

[0145] Step 14.3, according to the first linguistic variable membership set, the second linguistic variable membership set, and the pre-defined fuzzy control rule, a plurality of control instruction error fuzzy sets are obtained.

[0146] In the embodiment of the present application, the fuzzy rule of the heading angle control instruction error is shown in the following table:

[0147]

[0148] In the embodiment of the present application, the fuzzy rule of the speed control instruction error is shown in the following table:

[0149]

[0150] It should be noted that in the embodiments of the present application, only 7 linguistic variable values are used to describe the state error and the control instruction error, and therefore the speed control instruction error of the ith follower and the speed control instruction error correspond to 49 fuzzy sets respectively. Those skilled in the art should appreciate that if N linguistic variable values are used to describe the state error and the control instruction error, then each control instruction error corresponds to N 2 fuzzy sets.

[0151] Step 14.4, aggregate the plurality of control instruction error fuzzy sets to obtain a control instruction error aggregated fuzzy set.

[0152] Step 14.5, select the element set corresponding to the maximum membership degree from the control instruction error aggregated fuzzy set as a maximum control instruction error fuzzy set, and take the linguistic variable value corresponding to the maximum control instruction error fuzzy set as the optimal linguistic variable.

[0153] Step 14.6, determine the boundary of the follower control instruction fuzzy constraint according to the optimal linguistic variable to obtain the control instruction fuzzy constraint of the follower.

[0154] Step 14.6.1, for the ith follower, take the control instruction corresponding to the same membership degree of the optimal linguistic variable as the left boundary of the control instruction fuzzy constraint of the follower

[0155] Step 14.6.2, for the ith follower, take the control instruction corresponding to the same membership degree of the optimal linguistic variable as the right boundary of the control instruction fuzzy constraint of the follower

[0156] Step 14.6.3, determine the control instruction fuzzy constraint of the ith follower according to the left boundary and the right boundary

[0157] wherein U X represents the instruction space.

[0158] The specific process of step 15 (determining the control instruction of the follower according to the control instruction fuzzy constraint of the follower based on the distributed model predictive control method) will be described by way of example.

[0159] Step 15.1, calculate the control instruction of the ith follower according to the formula

[0160]

[0161] ​Obtain the coordinates of the i-th wingman relative to the j-th wingman in the reference point coordinate system X. R OY R State error in

[0162] Where j = 1, 2, ..., N a And j≠i. Here, the j-th wingman refers to the adjacent wingman of the i-th wingman.

[0163] Step 15.2, using the calculation formula

[0164]

[0165]

[0166] Xe ij （R） (τ|t k )=Xe j (τ|t k )-Xe i (τ|t k )

[0167]

[0168]

[0169] Obtain the control command trajectory of the i-th wingman. And the wingman in t k Control commands at +δ time This is determined to be the final control command.

[0170] Where δ represents the update period, t k This represents the time corresponding to the actual state quantity after k updates. Let τ represent the set of natural numbers, τ∈{t} k +δs}, s=0, 1, K, N p -1, N p δN represents the number of prediction steps. p S represents the prediction time domain. (X) and S (U) All represent normalized matrices, M1, M2, and N all represent positive definite symmetric matrices, and X represents a normalized matrix. i Let X represent the state variable of the i-th wingman. -i X represents the state variable of the wingman adjacent to the i-th wingman. i (t k ) indicates that the i-th wingman is in t k The state quantity at time U i (t k ) indicates that the i-th wingman is in tk the control variable at time t, denotes the predicted state trajectory of the ith wingman, denotes the estimated state trajectory of the ith wingman in the prediction horizon, denotes the optimal state trajectory of the ith wingman in the prediction horizon, denotes the predicted control trajectory of the ith wingman, denotes the estimated control trajectory of the ith wingman, denotes the optimal predicted control trajectory of the ith wingman, denotes the state transition equation of the ith wingman from time t to t + δ, denotes the predicted state of the ith wingman at time t + δ, denotes the estimated state of the ith wingman at time t + δ, denotes the optimal predicted state of the ith wingman at time t + δ, denotes the optimal predicted state of the neighboring wingman of the ith wingman at time t + δ, denotes the state error of the wingman i relative to the lead aircraft at time t, k denotes the coordinate transformation matrix from the inertial frame to the reference frame at time t, i denotes the state of the ith wingman at time t, k denotes the state of the ith wingman at time t, denotes the state of the target position of the ith wingman at time t, denotes the control of the target position of the ith wingman at time t, denotes the state error of the jth wingman relative to the lead aircraft at time t, denotes the estimated state of the neighboring wingman of the ith wingman at time t, k denotes the estimated state of the neighboring wingman of the ith wingman at time t, denotes the estimated control of the neighboring wingman of the ith wingman at time t, k denotes the state error of the ith wingman relative to the jth wingman at time t, denotes the rank of denotes the rank of denotes the control error of the wingman i relative to the lead aircraft at time t, ||·|| denotes the norm.

[0171] The analysis process of step 15 is as follows:

[0172] In the embodiments of the present application, when solving the distributed optimal control problem with the DMPC method, given the update period (or sampling period) δ and the prediction horizon T = δN P (N P denotes the number of steps of prediction), for the convenience of calculation, it can be considered that the distributed optimal control problem can be solved at time t k ​=The instantaneous synchronization at time t0+δk is resolved (t0 represents the initial time, t k This represents the time after k updates by instructions, where k ∈ N, and N represents the set of natural numbers.

[0173] It should be noted that common drone formations generally include multiple wingmen. Here, we will take a drone formation with multiple wingmen as an example to explain the method for determining wingman control commands in the drone formation provided in this application.

[0174] For distributed model prediction optimal control problems, it is necessary to consider the state influence of neighboring wingmen, that is, to consider the state variables X of each UAV's neighbors within each prediction domain. -i Estimation is performed. Specifically, the i-th wingman can receive the estimated control trajectory of its neighbor j from the previous moment, and the i-th wingman can also transmit its own estimated control trajectory to its neighbor j.

[0175] According to the i-th wingman at t k The state variable X at time t is i (t k ) and control quantity U i (t k ), which allows us to obtain the i-th wingman from t. k Predictive control trajectory starting at time (Indicates the control that can be exerted on the i-th wingman), the i-th wingman is in t k Optimal predictive control trajectory at time 1 (represents the predictive control trajectory that optimizes the objective function) and the i-th wingman at t k Estimated control trajectory at time (represents the estimate of the control exerted by adjacent wingmen on the i-th wingman), where the variable trajectory is a sequence of variable values ​​corresponding to different times τ in the prediction time domain, τ∈{t k +δs}, s=0, 1, K, N p -1. When the i-th wingman is in τ = t k At the initial state at time ,

[0176] The state change of the i-th wingman in the prediction time domain can be described by the following expression:

[0177]

[0178] Similarly, the predicted state trajectory of the i-th wingman can be obtained. Estimating the state trajectory and the optimal predicted state trajectory At τ=t k At that time, the value of each of the above trajectories is X.i (t k )。

[0179] From the above process, the estimated control trajectory of the adjacent wingman of the ith wingman can be defined as The estimated state trajectory of the adjacent wingman of the ith wingman is

[0180] According to the state error model of the ith wingman, the ith wingman and the leader j * The state error at time τ The ith wingman and the leader j * The control error at time τ The error trajectory of the ith wingman and the adjacent wingman j at time τ

[0181] Thus, the objective function of all wingman control and state variables can be obtained, which is as follows:

[0182]

[0183] where S (X) and S (U) both represent normalized matrices, which can normalize the state variable error and the control variable error to the range of [-1, 1], and M1, M2, N are all positive definite symmetric matrices, which are used to ensure the stability of the DMPC algorithm and determine the weights corresponding to different variables.

[0184] In combination with the fuzzy constraint of the ith wingman control instruction, the expression of the control instruction of the ith wingman at time t k is as follows:

[0185]

[0186] where k represents any given constant, x is the state space, U X is the instruction space, and is the fuzzy constraint space of the ith wingman.

[0187] Although the traditional DMPC algorithm can reduce the effect of the control variable by increasing the weight of the control variable, this will result in a longer convergence time. However, by using the DMPC control algorithm with fuzzy constraints, the amplitude change of the control variable can be reduced and the control variable can be kept constant for a period of time without changing the weight of the control variable, so as to reduce the control variable and reduce energy consumption.

[0188] In order to verify the effectiveness of the method for determining the wingman control instruction in the unmanned aerial vehicle formation provided in the present application, the method is simulated in the embodiments of the present application, and the process is as follows:​

[0189] In the simulation, a UAV formation consisting of six UAVs is considered, in which one UAV is the leader (UAV0) and the other five UAVs are the followers (UAV1, UAV2, UAV3, UAV4, UAV5). The initial values of the position coordinates, heading angles, velocities, and reference position coordinates of each UAV relative to the reference point are shown in the following table:

[0190] Serial number Position (m) Velocity (m / s) Heading angle (rad) Desired position (m) UAV0 (0，0) 35 pi / 2 - UAV1 (0，230) 35 pi / 2 (0，200) UAV2 (240，10) 35 pi / 2 (190，61) UAV3 (-340，150) 35 pi / 2 (-190，61) UAV4 (40，180) 35 pi / 2 (117，-161) UAV5 (-320，-260) 35 pi / 2 (-117，-161)

[0191] For the single UAV model, the autopilot parameters τ of the UAV v = 1 s, The state constraints V of the UAV min = 15 m / s, V max = 45 m / s, a max = 7 m / s, ω max = 0.1 rad / s, α max = 0.1 rad / s. For the simulation parameters of the UAV formation flight control, the prediction control domain N of the model predictive control p = 15, M1 = 2M2 = diag[15, 15, 5, 1, 0], N = diag[5, 1]; the variation range of the membership function in the fuzzy control, the input quantities including the position error, the heading angle error, and the velocity error are -500 ~ 500 (m), -pi / 2 ~ pi / 2 (rad), -20 ~ 20 (m / s), respectively, and the output quantities are -pi / 6 ~ pi / 6 (rad), -15 ~ 15 (m / s), respectively. The normalization matrix S is set as (X) = diag[500, 500, pi / 2, 20, 1] -1 and S (U) = diag[pi / 2, 20] -1 . The simulation time is Time = 90 s, and the simulation step is step = 0.5 s.

[0192] The simulation experiment includes the comparison of the DMPC method and the fuzzy DMPC method, and the comparison of the Latin hypercube sampling.

[0193] (1) Comparison of the DMPC method and the fuzzy constrained DMPC method.

[0194] In this simulation experiment embodiment, the leader is responsible for tracking the trajectory, and the followers fly according to the formation control algorithm, with the purpose of seeking the differences in energy consumption between the DMPC method and the fuzzy constrained DMPC method.

[0195] When the DMPC method and the fuzzy constrained DMPC method are used to control the flight of the followers, the flight trajectories of the followers are respectively as Figure 5a andFigure 5b Fig. 1 shows the simulation results of the DMPC method, and Fig. 2 shows the simulation results of the fuzzy-constrained DMPC method. The horizontal axis of both figures represents the east direction in the inertial coordinate system, and the vertical axis represents the north direction. It can be observed that each wingman constantly adjusts its position to form the desired formation. By comparing Fig. 1 and Fig. 2, it can be found that the trajectory generated by the DMPC method produces larger fluctuations, while the fuzzy-constrained DMPC method is more stable during the formation of the flight formation, and the flight trajectory is smoother. Figure 5a Figure 5b Figure 5a Figure 5b

[0196] To more intuitively show the simulation results, the state changes of UAV5, which has a larger position error, are selected as the observation object. The state trajectory of UAV5 under the control of the two methods (DMPC method and fuzzy-constrained DMPC method) is shown in Fig. 3 and Fig. 4, respectively. The horizontal axis of both figures represents the unit time, and the vertical axis represents the x-direction distance. Figure 6a Figure 6b Figure 6c Figure 6d Figure 6a Figure 6b Figure 6c Figure 6d

[0197] To better compare the DMPC method and the fuzzy-constrained DMPC method, the following four comparison indicators are defined:

[0198] 1. Flight distance I S Flight distance represents the total distance flown by the UAV from the beginning of the formation to the end.

[0199] 2. Speed change cumulative value I V The speed change cumulative value represents the cumulative absolute value of the speed change amount.

[0200] 3. Angle change cumulative value I Ψ The angle change cumulative value represents the cumulative absolute value of the angle change amount.

[0201] 4. Energy consumption I Qcsm Energy consumption represents the work done to overcome resistance. The resistance is proportional to the square of the speed, i.e. F drag = K drag y 2 For ease of calculation, the resistance coefficient K drag is set to 1. ​​​​​​​​​​​​​​​​

[0202] By calculation, the flight distance, the cumulative value of speed change and the cumulative value of angle change of DMPC method are 3266.75 m, 20.92 m / s and 0.71 rad respectively, while those of DMPC method with fuzzy constraint are 3263.76 m, 12.03 m / s and 0.48 rad respectively, and the changes of DMPC method with fuzzy constraint are 99.9%, 57.5% and 67.6% of those of DMPC method. The changes of energy consumption of DMPC method and DMPC method with fuzzy constraint are shown in Figure 7 , Figure 7 The abscissa represents unit time and the ordinate represents energy, and it can be seen from Figure 7 that the process of overcoming resistance work of DMPC method with fuzzy constraint is smooth. At the beginning, the speed of DMPC method increases rapidly, and more resistance work is done, showing an upward trend. Due to the large amplitude of speed increase and decrease of DMPC method, the speed reaches the same at about 18s, and then the speed of DMPC method is less than that of DMPC method with fuzzy constraint, resulting in less resistance work of DMPC method with fuzzy constraint, and the difference curve shows a downward trend, and finally tends to be stable. When the energy consumption difference is stable, the value is greater than zero, indicating that DMPC method with fuzzy constraint has the advantage of saving energy.

[0203] (II) Comparison of Latin hypercube sampling.

[0204] In order to further compare DMPC method with fuzzy constraint and DMPC method, Latin hypercube sampling (LHS) method (a method of approximately random sampling from a multi-parameter distribution, belonging to stratified sampling technique, commonly used in computer experiment or Monte Carlo integration, etc.) is used to obtain the initial position of UAV, and the target point in the formation topology is tracked from different initial positions. The differences in flight distance, cumulative value of speed change and cumulative value of angle change and energy consumption of DMPC algorithm and DMPC algorithm with fuzzy constraint are compared.

[0205] The leader flies along the Pe direction from the origin, and the target point G of the formation wingman is on the vertex of the regular pentagon constructed with the leader as the center. The sampling rule is to sample 1000 times in the range of 500m left and right of each target point G as the center for each UAV, and complete 1000 times of formation flight test.

[0206] The comparison simulation results of DMPC method with fuzzy constraint and DMPC method are shown in Figure 8a , Figure 8b , Figure 8c and Figure 8dThe four indexes, i.e. the difference of flight distance, the difference of cumulative value of speed change, the difference of cumulative value of angle change and the difference of energy loss, are compared from two aspects of single wingman (taking UAV5 as an example) and UAV formation, and the difference value is the simulation result of DMPC algorithm minus that of fuzzy constraint DMPC algorithm.

[0207] The horizontal coordinate of Figure 8a , Figure 8b , Figure 8c and Figure 8d represents the UAV number, and the vertical coordinate represents the distance; Figure 8a The horizontal coordinate of Figure 8b represents the UAV number, and the vertical coordinate represents the speed; Figure 8c The horizontal coordinate of Figure 8d represents the UAV number, and the vertical coordinate represents the angle; The horizontal coordinate of

[0208] represents the UAV number, and the vertical coordinate represents the energy, and the index (the difference of flight distance, the difference of cumulative value of speed change, the difference of cumulative value of angle change and the difference of energy loss) value of the fuzzy constraint DMPC method is generally better than that of the DMPC method. If the two methods are compared from the perspective of formation, the fuzzy constraint DMPC method is more obvious.

[0209]

[0210]

[0211] Since the initial positions of the sample points are different but have the same initial speed and angle, only the standard deviation of the angle and speed change is considered. It can be seen from the above table that:

[0212] 1) In the statistics of formation flight, it can be seen that in each simulation experiment, the indexes of the fuzzy constraint DMPC algorithm are better than those of the DMPC algorithm;

[0213] 2) In the statistics of single machine flight, the sample points with a difference value less than 0 not only have a small absolute value but also have a small proportion, among which the flight distance and angle change are less than 7%, the flight speed is less than 21%, and the energy loss difference is less than 0.2%, and the minimum value is -42.23, which is much smaller than the average value;

[0214] 3) The standard deviation of the angle change and speed change of the fuzzy constraint DPMC is about half of that of the DMPC method.

[0215] Statistical results show that the DMPC algorithm with fuzzy constraints is effective in reducing flight distance, cumulative angle change, and cumulative velocity change. Furthermore, the standard deviation indicates that the cumulative velocity change and cumulative angle change of the fuzzy-constrained DMPC method fluctuate relatively smoothly over the entire sample space.

[0216] Therefore, the method for determining wingman control commands in UAV formations provided in this application has the following advantages:

[0217] 1) The state error space of the wingman relative to the leader in the UAV formation is divided into fuzzy sets, and fuzzy constraints of control commands adapted to the error state of the wingman are designed according to the defined fuzzy rules. These constraints are used as control quantities constraints for the DMPC controller, which reduces the variation of speed and heading angle and reduces the energy consumption of UAVs in the formation process.

[0218] 2) Statistical results of the Latin supersquare sampling comparison simulation show that the method for determining wingman control commands in UAV formations provided in this application has a 99.8% and 100% probability of saving energy in single-aircraft and formation (6 aircraft) flight, respectively. From a statistical perspective, this method has good performance advantages in the entire sample space.

[0219] like Figure 9 As shown, embodiments of this application provide a terminal device, such as... Figure 9 As shown, the terminal device D10 of this embodiment includes: at least one processor D100 ( Figure 9 The diagram shows only one processor, a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, wherein the processor D100 executes the computer program D102 to implement the steps in any of the above method embodiments.

[0220] Specifically, the processor D100 executes the computer program D102, and according to a motion model of the unmanned aerial vehicle formation flight, a predefined leader initial state quantity and a predefined wingman initial state quantity, obtains an intermediate state quantity of the leader and an intermediate state quantity of the wingman under the action of the control quantity, then calculates a state error of the wingman relative to the leader according to the intermediate state quantity of the leader and the intermediate state quantity of the wingman, divides the predefined state error domain of the wingman and the predefined control instruction error domain by language variables, obtains a state error language variable set of the wingman and a control instruction error language variable set of the wingman, then determines a control instruction fuzzy constraint of the wingman according to a membership function corresponding to each language variable value in the state error language variable set, a membership function corresponding to each language variable value in the control instruction error language variable set and a predefined control instruction fuzzy control rule, and finally determines the control instruction of the wingman based on the distributed model predictive control method according to the control instruction fuzzy constraint of the wingman. Wherein, by determining the control instruction fuzzy constraint of the wingman and then determining the control instruction of the wingman based on the distributed model predictive control method, the amplitude change of the control quantity can be reduced under the condition that the control quantity weight is unchanged, so as to reduce the energy consumption of the wingman control.

[0221] The processor D100 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or can also be any conventional processor.

[0222] The storage D101 can be an internal storage unit of the terminal device D10, such as a hard disk or a memory of the terminal device D10, in some embodiments. The storage D101 can also be an external storage device of the terminal device D10, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device D10, in other embodiments. Further, the storage D101 can include both an internal storage unit and an external storage device of the terminal device D10. The storage D101 is used to store an operating system, an application program, a boot loader, data, and other programs, such as program codes of the computer program, etc. The storage D101 can also be used to temporarily store data that has been output or will be output.

[0223] It should be noted that the information interaction and execution process between the above apparatuses / units are based on the same concept as the method embodiments of the present application, and the specific functions and technical effects brought by the above apparatuses / units can be referred to the method embodiments part, which will not be repeated here.

[0224] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional units and modules is taken as an example for illustration, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules to complete all or part of the above described functions. Each functional unit and module in the embodiments can be integrated in one processing unit, or each unit can exist physically independently, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for convenient distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the system can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0225] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the present application can implement all or part of the processes in the above-mentioned embodiment methods through a computer program to instruct relevant hardware to complete, and the computer program can be stored in a computer readable storage medium. When the computer program is executed by a processor, the steps of each method embodiment described above can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files or some intermediate forms. The computer readable medium at least includes any entity or device capable of carrying the computer program code to the terminal device described above, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium. For example, U disk, mobile hard disk, magnetic disk or optical disk, etc. In some jurisdictions, according to legislation and patent practice, the computer readable medium can not be an electrical carrier signal and a telecommunication signal.

[0226] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0227] Those of ordinary skill in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0228] In the embodiments provided by the present application, it should be understood that the disclosed apparatus / network device and method can be implemented in other ways. For example, the above-described apparatus / network device embodiments are merely schematic, for example, the division of the modules or units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0229] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may also be distributed to multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiment of the present application according to actual needs.

[0230] The above is the preferred embodiment of the present application. It should be pointed out that for ordinary skilled persons in the art, without departing from the principles described in the present application, a number of improvements and refinements can be made, which should also be considered as the protection scope of the present application.

Claims

1. A method for determining control instructions of a wingman in a UAV formation, characterized in that, The application relates to a method for controlling a UAV formation flight, and belongs to the field of UAV formation flight control. According to a motion model of the UAV formation flight, pre-defined initial state quantities of a long plane and pre-defined initial state quantities of a wing plane, intermediate state quantities of the long plane and intermediate state quantities of the wing plane are obtained under the action of control quantities; According to the intermediate state quantities of the long plane and the intermediate state quantities of the wing plane, state error of the wing plane relative to the long plane is calculated; Language variable division is performed on a pre-defined state error domain of the wing plane and a pre-defined control instruction error domain, so as to obtain a state error language variable set of the wing plane and a control instruction error language variable set of the wing plane; According to a membership function corresponding to each language variable value in the state error language variable set, a membership function corresponding to each language variable value in the control instruction error language variable set and pre-defined control instruction fuzzy control rules, control instruction fuzzy constraints of the wing plane are determined; According to the control instruction fuzzy constraints of the wing plane, control instruction of the wing plane is determined based on a distributed model prediction control method; According to the control instruction fuzzy constraints of the wing plane, control instruction of the wing plane is determined based on the distributed model prediction control method, and the method comprises the following steps: a calculation formula is used ； Get the first The wingman was relative to the first The wingman is in the reference point coordinate system State error in ;in, ; a calculation formula is used ； ； ； Get the first The control command trajectory of the wingman and the wingman in Time control commands This is determined to be the final control command; among which, among which, Indicates the update cycle. Indicates the process The time corresponding to the actual state quantity after the next update. , Represents the set of natural numbers. , Indicates the number of prediction steps. Indicates the prediction time domain, and Both represent normalized matrices. Both represent positive definite symmetric matrices. Indicates the first Status parameters of each wingman. Indicates the relationship with the first The state variables of adjacent wingmen. Indicates the first A wingman The state quantity at any given time. Indicates the first A wingman The amount of control at any given moment Indicates the first The predicted state trajectory of each wingman in the prediction time domain. Indicates the first The estimated state trajectory of each wingman in the prediction time domain. Indicates the first The optimal state trajectory of each wingman in the prediction time domain. Indicates the first Predictive control trajectory of a wingman Indicates the first One wingman estimates the control trajectory. Indicates the first The optimal predictive control trajectory of a wingman. Indicates the first A wingman Time's up State transition equation at time t. Indicates the first A wingman The predicted state quantity at time t, Indicates the first A wingman The estimated state variables at time t. Indicates the first A wingman The optimal predicted state quantity at time 1. optimal predicted state of the neighboring wingman of the i-th wingman at time k, norm.​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ 2. The determination method according to claim 1, characterized in that, The expression of the motion model of the UAV formation flight is as follows: ； ； ； in, Indicates the first in the aforementioned drone formation The coordinates of the wingman in the inertial coordinate system. Indicates the first The wingman's heading angle Indicates the first The speed of the wingman Indicates the first The angular velocity of the wingman Indicates the first Heading angle control commands for each wingman Indicates the first Speed ​​control commands for the wingman and Both represent the inertial time constant of the autopilot's heading angle channel. The inertial time constant of the velocity channel. , The total number of wingmen in the flight formation. This indicates the minimum flight speed of the wingman. This indicates the wingman's maximum flight speed. This indicates the wingman's maximum acceleration. This indicates the wingman's maximum angular velocity. This indicates the wingman's maximum angular acceleration. Indicates the first Angular acceleration of the wingman Indicates the first The acceleration of the wingman.

3. The determination method according to claim 2, characterized in that, the intermediate state quantity of the lead aircraft ; the intermediate state quantity of the wingman ; wherein denotes the intermediate state quantity of the th wingman According to the intermediate state quantities of the long plane and the intermediate state quantities of the wing plane, state error of the wing plane relative to the long plane is calculated, and the method comprises the following steps: a north-east coordinate system as an inertial coordinate system and taking the long machine as a reference point , obtaining a position of the reference point a state quantity in the inertial coordinate system , ; the position of the reference point and the state quantity of the reference point are determined by a state quantity of the long machine; establishing a reference point coordinate system with the lead aircraft as a reference point in the inertial coordinate system , and determining a target point of the i th wingman in the reference point coordinate system ; and ​ a calculation formula is used ； Obtaining the first subordinate machine in the inertial coordinate system target point state quantity , wherein , , ; a calculation formula is used ； obtaining the state error , denotes the state error of the th wingman relative to the lead aircraft in the reference point coordinate system; wherein, denotes the coordinate transformation matrix from the inertial frame to the reference frame, , , , and each denotes a state quantity of the th wingman in the reference point coordinate system, , denotes the error of the th wingman relative to the lead aircraft in the reference point coordinate system in the direction, denotes the error of the th wingman relative to the lead aircraft in the reference point coordinate system in the direction, denotes the heading angle error of the th wingman relative to the lead aircraft in the reference point coordinate system, denotes the velocity error of the th wingman relative to the lead aircraft in the reference point coordinate system, denotes the angular velocity error of the th wingman relative to the lead aircraft in the reference point coordinate system.

4. The determination method according to claim 3, characterized in that, The control instruction error includes a speed control instruction error of the first wingman and a heading angle control instruction error of the first wingman .

5. The determination method according to claim 4, characterized in that, The language variable comprises NB, NM, MS, Z, PS, PM and PB; wherein NB represents a relatively large negative error, NM represents a medium negative error, MS represents a relatively small negative error, Z represents almost no error, PS represents a relatively small positive error, PM represents a medium positive error and PB represents a relatively large positive error; the state error linguistic variable set , denotes the state error linguistic variable set of the ith wingman; wherein, , , , ; the control command error language variable set , represents the control command error language variable set of the th wingman; wherein , .

6. The determination method according to claim 5, characterized in that, According to a membership function corresponding to each language variable value in the state error language variable set, a membership function corresponding to each language variable value in the control instruction error language variable set and pre-defined control instruction fuzzy control rules, control instruction fuzzy constraints of the wing plane are determined, and the method comprises the following steps: According to the membership function corresponding to each language variable value in the state error language variable set, a state error membership degree image is drawn, so as to obtain a first language variable membership degree set corresponding to the state error; the first language variable membership degree set comprises the membership degrees corresponding to each language variable value in the state error language variable set; According to the membership function corresponding to each language variable value in the control instruction error language variable set, a control instruction error membership degree image is drawn, so as to obtain a second language variable membership degree value corresponding to the control instruction error; the second language variable membership degree set comprises the membership degrees corresponding to each language variable value in the control instruction error language variable set; According to the first language variable membership degree set, the second language variable membership degree set and pre-defined fuzzy control rules, a plurality of control instruction error fuzzy sets are obtained; The plurality of control instruction error fuzzy sets are aggregated, so as to obtain a control instruction error aggregated fuzzy set; and The plurality of control instruction error fuzzy sets are aggregated, so as to obtain a control instruction error aggregated fuzzy set; and From the control instruction error aggregated fuzzy set, an element set corresponding to a maximum membership degree is selected as a maximum control instruction error fuzzy set, and a language variable value corresponding to the maximum control instruction error fuzzy set is taken as an optimal language variable; A boundary of a control instruction fuzzy constraint of the wingman is determined according to the optimal language variable, to obtain a control instruction fuzzy constraint of the wingman.

7. The determination method according to claim 6, characterized in that, The control instruction fuzzy control rule comprises: According to and determining a speed control command error for the nth wingman According to and determine the heading angle control instruction error of the first wingman .

8. The determination method according to claim 7, characterized in that, The control instruction fuzzy control rule comprises: For the first subordinate, the control instruction corresponding to the same membership degree of the optimal language variable of the language variable value adjacent to the left side of the optimal language variable is taken as the left boundary of the subordinate control instruction fuzzy constraint ; For the first subordinate, the control instruction corresponding to the same membership degree of the optimal language variable of the language variable value adjacent to the right side of the optimal language variable is taken as the right boundary of the subordinate control instruction fuzzy constraint ; According to the left boundary and the right boundary Determine the first Fuzzy constraints on the control commands of individual wingmen ,in, Indicates the instruction space.

9. A terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The control instruction fuzzy control rule comprises: The processor implements the method for determining a control instruction of a wingman in a UAV formation according to any one of claims 1 to 8 when executing the computer program.

Images