Unmanned aerial vehicle group specified time distributed resource scheduling method based on time-varying gain

Abstract

The invention discloses an unmanned aerial vehicle group specified time distributed resource scheduling method based on time-varying gain, which belongs to the field of unmanned aerial vehicle group scheduling, and comprises the following steps: S1, constructing an unmanned aerial vehicle group resource scheduling model and a communication topology model with global equality constraints according to a multi-agent system, and obtaining an unmanned aerial vehicle group state; s2, constructing an unmanned aerial vehicle group state updating model based on the derivative information of the unmanned aerial vehicle group state; s3, designing time-varying gain parameters in the unmanned aerial vehicle group state updating model constructed in the step S2; and S4, substituting the time-varying gain parameter designed in the step S3 into the step S2, and updating the constructed unmanned aerial vehicle group state updating model. By adopting the time-varying gain-based distributed resource scheduling method for the unmanned aerial vehicle group at the specified time, additional auxiliary variables are not needed, the communication consumption of the system is reduced, and the optimal solution of the problem can be converged within the pre-specified time, so that the requirement on the real-time performance of an algorithm in practical application is met.

Description

Technical Field

[0001] The present invention relates to the field of UAV swarm scheduling, and particularly to a method for distributed resource scheduling of UAV swarms at a specified time based on time-varying gain. Background Art

[0002] Large-scale UAV swarms have the characteristic of distributed communication. The purpose of distributed optimization of UAV swarms is to design a distributed algorithm to find the optimal solution of the global optimization problem through communication, exchange, and cooperation among UAV swarms.

[0003] Compared with traditional centralized optimization algorithms, distributed optimization algorithms have the following obvious advantages: Each UAV swarm only needs to exchange information with its neighbors. In this way, not only can the security of the network system be improved, but also the cost of communication can be reduced, so that the information and objective function of each UAV swarm can be effectively protected. Therefore, it is of great significance to study the distributed optimization problem of UAV swarm systems.

[0004] Although significant progress has been made in the distributed continuous-time optimization of UAV swarms, most of the existing distributed resource scheduling algorithms are exponentially convergent. To cope with uncertain factors and interference in practical applications, UAV swarms are often required to respond quickly. Therefore, it is very necessary to study a distributed resource scheduling algorithm for UAV swarms that can converge to the optimal solution at a specified time.

[0005] Currently, there are few existing distributed resource scheduling algorithms that can achieve convergence at a specified time under a directed graph. Mainly, a fixed-time convergent undirected graph distributed continuous-time scheduling algorithm is realized based on the gradient descent method by introducing auxiliary variables (such as "A fixed-time convergent algorithm for distributed convex optimization in multi-agent systems" published by Chen, Gang, and Zhiyong Li). However, the convergence time of the fixed-time related algorithm cannot be changed, and it is relatively limited in practical application scenarios. For this reason, research on specified-time convergence based on time-varying gain has been carried out (such as "Prescribed-time consensus and containment control of networked multiagent systems" published by Wang, Yujuan, etc.). However, the existing research has not introduced the specified-time convergence method based on time-varying gain into distributed resource scheduling. Summary of the Invention

[0006] To solve the problem of distributed resource scheduling for unmanned aerial vehicle (UAV) swarms under equality constraints in the existing technology, the present invention provides a method for distributed resource scheduling of UAV swarms at a specified time based on time-varying gain, which does not require the introduction of additional auxiliary variables, reduces the communication consumption of the system, and can converge to the optimal solution of the problem within a pre-specified time to meet the real-time requirements of the algorithm for practical applications.

[0007] The method for distributed resource scheduling of UAV swarms at a specified time based on time-varying gain according to the present invention includes the following steps: S1. Construct a UAV swarm resource scheduling model and a communication topology model with global equality constraints according to the multi-agent system, and obtain the UAV swarm state; S2. Based on the derivative information of the UAV swarm state obtained in step S1, construct a UAV swarm state update model; S3. Design the time-varying gain parameter in the UAV swarm state update model constructed in step S2; S4. Substitute the time-varying gain parameter designed in step S3 into step S2 to update the constructed UAV swarm state update model.

[0008] Preferably, the formula of the UAV swarm resource scheduling model constructed in step S1 is as follows: (1) In the formula, represents the number of UAV swarms; represents the state of the UAV swarm and represents the proportion of the number of UAVs allocated to the UAV swarm in the total number of UAVs; represents the resource allocation situation of each UAV swarm, and satisfies the equality constraint ; represents the cost function of the UAV swarm which satisfies the strong convexity condition , where is a positive constant, represents the set of real numbers, is an arbitrary real number, represents the second derivative of

[0009] Preferably, the communication topology model constructed in step S1 is defined as follows: The communication topology model of the UAV swarm is regarded as a multi-agent network system composed of nodes, and the communication topology structure is denoted as , where represents the node set, and ; the edge set Represents a node Receiving node receives the information of, and , ; At the same time, it is set that the path from the slave node to the node is an ordered sequence of edges , ; The set composed of the inner neighbors of the node is .

[0010] Preferably, for the communication topology graph described in step S1 , where represents the element in the th row and th column of the matrix , and it is defined that , then , otherwise .

[0011] Preferably, for the communication topology graph described in step S1 , and it is defined that , when , , where represents the element in the th row and th column of the Laplacian matrix , represents the element in the th row and th column of the Laplacian matrix ; Preferably, the communication topology model of the UAV swarm described in step S1 meets the following requirements: The communication topology graph is a strongly connected balanced graph, for , and it satisfies , , represents the left multiplication of the Laplacian

[0012] by (2) In the formula, represents the derivative of the state of the UAV swarm ; is a constant greater than 0; is a parameter to be designed; is a time-varying gain to be designed; is the derivative of the time-varying gain; is the state of the UAV swarm at time ; is the state of the UAV swarm at time .

[0013] Preferably, in step S3, the time-varying gain is calculated as follows: (3) The derivative of the time-varying gain is calculated as follows: (4) The parameter to be designed is calculated as follows: (5) In the formula, represents the initial time; represents the pre-specified convergence time; is a constant greater than 2; represents the two-norm of the Laplacian matrix; represents the smallest non-zero eigenvalue of the Laplacian; represents the smallest non-zero eigenvalue of

[0014] The present invention has the following beneficial effects: (1) Utilize derivative information to design a distributed algorithm, and only the information of itself and its neighbors is required during the algorithm iteration; (2) Design a time-varying gain that can achieve convergence at a specified time, and update the algorithm based on this time-varying gain to meet the real-time requirements of the algorithm in practical applications; (3) Do not need to introduce additional auxiliary variables, reducing the communication consumption of the system; (4) Design the parameter conditions for the algorithm to converge.

[0015] Next, through the accompanying drawings and embodiments, the technical solution of the present invention will be further described in detail. Description of the Drawings

[0016] Figure 1 is the flowchart of the method for specifying time distributed resource scheduling of UAV swarms based on time-varying gain of the present invention; Figure 2 It is a schematic diagram of the communication topology model of the UAV swarm in the simulation experiment of the present invention; Figure 3 It is a schematic diagram of the resource ratio allocated to each UAV swarm in the simulation experiment of the present invention; Figure 4 It is a schematic diagram of the change in the total UAV resources in the simulation experiment of the present invention; Figure 5 It is a schematic diagram of the change in the total system cost in the simulation experiment of the present invention. Detailed implementation manners

[0017] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the following further details the embodiments of the present invention in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the embodiments of the present invention, and are not used to limit the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts belong to the scope protected by this application. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end.

[0018] It should be noted that the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0019] Similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0020] As Figure 1 shown, the time-varying gain-based UAV swarm specified-time distributed resource scheduling method of the present invention includes the following steps: S1. Construct a UAV swarm resource scheduling model and a communication topology model with global equality constraints according to the multi-agent system, and obtain the UAV swarm state; The formula of the UAV swarm resource scheduling model constructed in step S1 is as follows: (1) In the formula, represents the number of UAV swarms; represents the state of the UAV swarm , representing the proportion of the number of UAVs allocated to the UAV swarm in the total number of UAVs; It represents the resource allocation situation of each UAV group, and satisfies the equality constraint , that is, the sum of the resource allocations of all UAV groups is ; represents the cost function of UAV group , which satisfies the strong convexity condition , is a positive constant, represents the set of real numbers, is an arbitrary real number, represents the second derivative of with respect to

[0021] The communication topology model constructed in step S1 is defined as follows: The communication topology model of the UAV group is regarded as a multi-agent network system composed of nodes. The communication topology structure is denoted as , where represents the set of nodes, and ; the edge set means that node receives the information of receiving node , and , ; At the same time, it is set that the path from node to node is an ordered edge sequence , ; the set composed of the in-neighbors of node is .

[0022] The adjacency matrix of the communication topology graph described in step S1 is , where represents the -th row and -th column element of the matrix , and it is defined that ; otherwise The Laplacian matrix of the communication topology graph described in step S1 is , and it is defined that . When , , where represents the -th row and Denote the Laplacian matrix The row column element; The communication topology model of the UAV swarm described in step S1 satisfies the following requirements: The communication topology graph is a strongly connected balanced graph. For , And it satisfies , Denote the Laplacian Left multiply , Denote the transpose of a vector with all elements being 1.

[0023] S2. Based on the derivative information of the UAV swarm state obtained in step S1, construct a UAV swarm state update model; In step S2, the expression of the UAV swarm state update model is as follows: (2) In the formula, Denote the derivative of the UAV swarm state; Is a constant greater than 0; Is a parameter to be designed; Is a time-varying gain to be designed; Is the derivative of the time-varying gain; Is The state of the UAV swarm at time ; Is The state of the UAV swarm at time ;

[0024] S3. Design the time-varying gain parameter in the UAV swarm state update model constructed in step S2.

[0025] In step S3, the calculation formula of the time-varying gain is as follows: (3) The derivative of the time-varying gain The calculation formula is as follows: (4) The parameter to be designed (5) In the formula, Denote the initial time; Denote the pre-specified convergence time; Is a constant greater than 2; Denotes the two - norm of the Laplacian matrix; Denotes the Laplacian of the smallest non - zero eigenvalue; Denotes the smallest non - zero eigenvalue of.

[0026] S4. Substitute the time - varying gain parameters designed in step S3 into step S2 to update the constructed UAV swarm state update model.

[0027] Simulation experiment In this simulation experiment, construct a UAV swarm communication topology model as Figure 2 shown. Considering that three UAV swarms need to execute their respective target tasks, the cost function of UAV swarm 1 is , the cost function of UAV swarm 2 is , and the cost function of UAV swarm 3 is . As Figure 2 shown, this topology diagram satisfies the balanced graph and is strongly connected. Set the initial values of each UAV swarm to , and , and set , , , . Each UAV swarm is updated according to the algorithm given in the present invention. The results are as Figures 3 - 5 shown, Figure 3 indicating that the states of each UAV swarm converge to the optimal solution of the distributed resource scheduling model under equality constraints at the specified time; Figure 4 indicating that during the algorithm iteration process, the total UAV resources remain unchanged, that is, the equality constraint holds in real time. Figure 5 indicating that the total system cost converges to the optimal solution at the specified time.

[0028] The overall experimental results show that each UAV swarm realizes resource allocation with the total resources unchanged and within the specified time. The algorithm converges at the specified time, proving that the method of UAV swarm specified - time distributed resource scheduling based on time - varying gain proposed in the present invention can meet the real - time requirements of the algorithm in practical applications.

[0029] Therefore, the present invention adopts the above - mentioned method of UAV swarm specified - time distributed resource scheduling based on time - varying gain, without additional auxiliary variables, reducing the communication consumption of the system, and being able to converge to the optimal solution of the problem within the pre - specified time to meet the real - time requirements of the algorithm in practical applications.

[0030] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions of the present invention or make equivalent replacements, and these modifications or equivalent replacements do not enable the modified technical solutions to depart from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for scheduling distributed resources at designated time for drone swarms based on time-varying gain, characterized by: The following steps are involved: S1. Construct a UAV swarm resource scheduling model and communication topology model with global equality constraints based on the multi-agent system to obtain the UAV swarm status; S2. Based on the derivative information of the drone swarm state obtained in step S1, a drone swarm state update model is constructed; S3, designing the time-varying gain parameters in the drone swarm state update model constructed in step S2; S4. Substitute the time-varying gain parameters designed in step S3 into step S2 to update the constructed drone swarm state update model.

2. The method for scheduling distributed resources of drone swarms at designated time based on time-varying gain according to claim 1 is characterized by: The formula of the drone swarm resource scheduling model constructed in step S1 is as follows: （1） In the formula, Indicates the number of drone swarms; Represents a drone swarm The state represents the drone swarm The number of assigned drones as a percentage of the total number of drones; represents the resource allocation of each drone group, and Satisfy equality constraints ; Represents a drone swarm The cost function satisfies the strong convexity condition , is a normal number, represents the set of real numbers, is any real number, express right The second derivative of .

3. The method for scheduling distributed resources of drone groups at designated time based on time-varying gain according to claim 2 is characterized by: The communication topology model constructed in step S1 is defined as follows: The communication topology model of the drone swarm is considered to be composed of The multi-agent network system consists of nodes, and the communication topology is recorded as ,in, represents a set of nodes, and ; Edge Set Representation Node Receiving Node information, and , ; Set the slave node at the same time To Node A path is an ordered sequence of edges , ;node The set of inner neighbors is .

4. The method for scheduling distributed resources of drone swarms at designated time based on time-varying gain according to claim 3 is characterized by: Communication topology diagram The adjacency matrix of ,in Representation Matrix No. OK List elements and define , at this time if ,but ,otherwise .

5. The method for scheduling distributed resources of drone swarms at designated time based on time-varying gain according to claim 4 is characterized by: The communication topology diagram described in step S1 The Laplace matrix of , and define ,when , ,in represents the Laplacian matrix No. OK List elements, represents the Laplacian matrix No. OK Column elements.

6. The method for scheduling distributed resources of drone groups at designated time based on time-varying gain according to claim 5 is characterized by: The communication topology model of the drone swarm described in step S1 meets the following requirements: Communication topology diagram is a strongly connected balanced graph. , And satisfy , Represents Laplace Left multiplication , Represents the transpose of a vector whose elements are all 1.

7. The method for scheduling distributed resources of drone groups at designated time based on time-varying gain according to claim 6 is characterized by: In step S2, the expression of the drone swarm state update model is as follows: （2） In the formula, Represents a drone swarm Derivatives of states; is a constant greater than 0; is the parameter to be designed; is the time-varying gain to be designed; is the derivative of the time-varying gain; yes Drone swarm Status; yes Drone swarm status.

8. The method for scheduling distributed resources of drone groups at designated time based on time-varying gain according to claim 7 is characterized by: In step S3, the time-varying gain The calculation formula is as follows: （3） Derivative of time-varying gain The calculation formula is as follows: （4） Parameters to be designed The calculation formula is as follows: （5） In the formula, represents the initial moment; Indicates the time for pre-specified convergence; is a constant greater than 2; Represents Laplace The second norm of a matrix; Represents Laplace The smallest non-zero eigenvalue of ; express The smallest non-zero eigenvalue of .

Images