CN-115935560-B - Multi-unmanned aerial vehicle network coverage optimization method

CN115935560BCN 115935560 BCN115935560 BCN 115935560BCN-115935560-B

Abstract

The invention provides a network coverage optimization method for multiple unmanned aerial vehicles, and belongs to the technical field of information. The optimization method comprises 1) searching candidate solutions in a balance pool by using a Lewy flight mode to generate a local population, calculating the fitness value of particles in the local population, replacing the particles in the original balance pool with particles with higher fitness value, 2) adopting a nonlinear decreasing strategy for the number of the particles in the population, eliminating particles with low fitness value in the iterative process, reducing the operation amount of the algorithm, and 3) applying the novel balance optimization algorithm to the network coverage problem of multiple unmanned aerial vehicles, and defining a group of position coordinate codes of unmanned aerial vehicles in a two-dimensional plane as one particle. Simulation experiments show that the multi-unmanned aerial vehicle network coverage optimization method provided by the invention better balances global search performance and local optimization performance, has higher convergence rate, has the function of planning spatial position distribution of isomorphic and heterogeneous multi-unmanned aerial vehicles, and effectively improves network coverage rate.

Inventors

GUO JIANDONG
DENG DEHUI
SUN XIAOYUAN
LIANG CHENYU
LIAO ZHIXIANG

Assignees

南京航空航天大学

Dates

Publication Date: 20260505
Application Date: 20221025

Claims (4)

1. The multi-unmanned aerial vehicle network coverage optimization method is characterized by comprising the following steps of: modeling a network coverage optimization problem of multiple unmanned aerial vehicles, encoding position coordinates of unmanned aerial vehicles in a two-dimensional plane into particles in an optimization algorithm, and setting the dimensions of single particles as follows The dimensions of the dimensions, The method comprises the steps of setting the boundary of a target area as the upper boundary and the lower boundary of a solution space, setting the network coverage rate of the target area as an objective function of an optimization algorithm; generating a local population for the candidate solutions by adopting a Lewy flight mode on the basis of a balance optimization algorithm, and updating and replacing the original candidate solutions in a balance pool according to fitness values, and dynamically adjusting the particle number in an iteration process by adopting a nonlinear decreasing population strategy; selecting four particles with optimal histories and average values thereof to form a balance pool: (2) Wherein, the , , , For the four particles to be historically optimal, Is the average state of four particles; And generating a local population by adopting a Lewy flight mode for the candidate solutions in the balance pool, wherein the generation mode is as follows: (10) In the formula, Four particles with optimal histories in the balancing pool are respectively corresponding, Is the average state of four particles in the original balancing tank, , For the number of desired local populations, Step length obtained by the Lewy flight mode: (11) (12) In order for the attenuation factor to be a factor, Is a constant value, and is used for the treatment of the skin, , Compliance with And Is used for the distribution of the gaussian distribution of (c), Is a gamma function; The updating rule of the new candidate solution is that the formula (10) respectively generates the local population of four particles with optimal histories, calculates the fitness value of the particles in the local population, compares the fitness value with the fitness value of the corresponding candidate solution, and replaces the corresponding candidate solution with the particles with better fitness value; in the iterative process, a strategy of nonlinearly decreasing the number of particles of a variant group is adopted, and the number of particles in the population in each iterative process is as follows: (13) In the formula, For the number of particles of the current iteration process, For the number of particles of the initial population, For the number of particles in the final population, For the current number of iterations, For the maximum number of iterations to be performed, After each iteration is finished, calculating the particle number of the population at the next iteration, sorting according to the fitness value, eliminating particles with low fitness value, and taking the rest particles as the initial population at the next iteration; and decoding the historical optimal particles solved by the algorithm to obtain the spatial position distribution and the network coverage area schematic diagram of the unmanned aerial vehicle group.
2. The multi-unmanned aerial vehicle network coverage optimization method of claim 1, wherein the specific process of solving and optimizing the multi-unmanned aerial vehicle network coverage problem is as follows: S1, setting the particle numbers of an initial population and a final population, and initializing parameters of an improved balance optimization algorithm; S2, setting upper and lower boundaries of a solution space and the number of unmanned aerial vehicles according to network coverage area information, generating initial population particle distribution, and initializing a balance pool; s3, calculating the fitness value of single particles in the population according to the evaluation function, comparing the fitness value with a balance Chi Zhongli, and updating a balance pool; s4, calculating an exponential term coefficient F and a generation rate G, and updating according to an algorithm updating formula; s5, generating local population particles for the candidate solutions in a Lewy flight mode, and updating the candidate solutions in the balance pool according to the fitness value; s6, calculating a particle fitness value, adopting a strategy of nonlinearly decreasing the number of particles of a variation group, eliminating the particles of a set number according to the fitness value, and calculating the number of the particles in the next iteration process; and S7, judging whether the maximum iteration times are reached, if so, outputting the optimal solution in the balance pool, and establishing the optimal solution as the spatial distribution of the unmanned aerial vehicle in the target area, otherwise, returning to the step S3.
3. The multi-unmanned aerial vehicle network coverage optimization method of claim 2, wherein the area ratio of the union of all unmanned aerial vehicle coverage areas to the target area is defined as network coverage, the network coverage is taken as an evaluation index, and an evaluation function is designed as follows: (17) Wherein, the Is the coverage area of a single unmanned aerial vehicle, The coverage rate is higher, and the evaluation function value is larger when the number of unmanned aerial vehicles is fixed.
4. The multi-unmanned aerial vehicle network coverage optimization method of claim 2 or 3, wherein the following four algorithms are selected for comparison with the multi-unmanned aerial vehicle network coverage optimization method, and the results are subjected to performance analysis: The balance Optimization algorithm Equilibrium Optimizer EO, the Cuckoo Search CS, the particle swarm Optimization algorithm PARTICLE SWARM Optimization PSO and the gray wolf Optimization algorithm Grey Wolf Optimizer GWO are set under the condition that the number of unmanned aerial vehicles is the same, the same iteration times are set, the coverage rate curve change graph and the final coverage rate are analyzed, and the analysis results of the Optimization method are obtained.

Description

Multi-unmanned aerial vehicle network coverage optimization method Technical Field The invention relates to the field of network coverage optimization of multiple unmanned aerial vehicles, in particular to design and application of a network coverage optimization method of multiple unmanned aerial vehicles. Background The realization of data relay by constructing a high-altitude base station through the unmanned aerial vehicle mounted communication equipment becomes a research hotspot in the engineering field. The coverage area of a single unmanned aerial vehicle is limited, and a mode of networking a plurality of unmanned aerial vehicles is adopted to realize a wider coverage area. When the traditional methods such as a virtual force field method and a geometric calculation method solve the problem of wireless network node deployment, the problems of low coverage rate, large calculated amount and the like exist, and the intelligent algorithm has good performance on the problems. However, when the existing intelligent algorithms solve the problem of network coverage optimization, the problems of low convergence speed, low coverage rate and the like exist, and the expected target cannot be achieved frequently. Therefore, it is particularly important to design and verify a feasible multi-unmanned-plane network coverage optimization method. Disclosure of Invention The invention provides a multi-unmanned aerial vehicle network coverage optimization method and a verification method which are good in performance and easy to realize, and aims to realize the multi-unmanned aerial vehicle network coverage deployment task. The technical scheme is as follows: The application provides a network coverage optimization method of multiple unmanned aerial vehicles, which comprises the following steps: modeling a network coverage optimization problem of multiple unmanned aerial vehicles, encoding position coordinates of unmanned aerial vehicles in a two-dimensional plane into particles in an optimization algorithm, and setting the dimensions of single particles as follows The dimensions of the dimensions,The method comprises the steps of setting the boundary of a target area as the upper boundary and the lower boundary of a solution space, setting the network coverage rate of the target area as an objective function of an optimization algorithm; Generating a local population for the candidate solutions by adopting a Lewy flight mode on the basis of a balance optimization algorithm, and updating and replacing the original candidate solutions in a balance pool according to fitness values, and dynamically adjusting the particle number in an iteration process by adopting a strategy of nonlinear decreasing the particle number of the variant group; and decoding the historical optimal particles solved by the algorithm to obtain the spatial position distribution and the network coverage area schematic diagram of the unmanned aerial vehicle group. Further, the specific process of solving and optimizing the network coverage problem of the multiple unmanned aerial vehicles is as follows: s1, setting initial population and final population particle number, and initializing parameters of an improved balance optimization algorithm; s2, setting upper and lower boundaries of a solution space according to network coverage area information, generating initial particle distribution according to the number of unmanned aerial vehicles, and initializing a balance pool; s3, calculating the fitness value of single particles in the population according to the evaluation function, comparing the fitness value with the balance pool, and updating the balance pool; s4, calculating an exponential term coefficient F and a generation rate G, and updating according to an algorithm updating formula; s5, generating a local population for the candidate solutions in a Lewy flight mode, and updating and replacing the original candidate solutions in the balance pool according to the fitness value; s6, calculating a particle fitness value, adopting a strategy of nonlinearly decreasing the number of particles of a variation group, eliminating the particles of a set number according to the fitness value, and calculating the number of the particles in the next iteration process; and S7, judging whether the maximum iteration times are reached, outputting the optimal solution in the balance pool if the maximum iteration times are reached, decoding the optimal solution into the spatial distribution of the unmanned aerial vehicle in the target area, and otherwise, returning to the step S3. Further, four particles with optimal histories and average values thereof are selected to form a balance pool: (2) Wherein, the ,,,For the four particles to be historically optimal,Is the average state of four particles; And generating a local population by adopting a Lewy flight mode for the candidate solutions in the balance pool, wherein the generation mode is as follows: (10) In th