﻿ 基于改进粒子群算法的高速载荷控制参数优化
 舰船科学技术  2024, Vol. 46 Issue (11): 58-62    DOI: 10.3404/j.issn.1672-7649.2024.11.011 PDF

High-speed load control parameter optimization based on improved particle swarm algorithm
FAN Liwei, FAN Hui, LV Jianguo, LV Rui, LIU Wei, LIN Longyan
The 705 Research Institute of CSSC, Xi′an 710077, China
Abstract: The instantaneous control of the synthetic angle of attack needs to be realized before the high-speed load enters the water, and the instantaneous attitude control parameter of the high-speed load is the key to determining whether the high-speed load can enter the water within a reasonable synthetic angle of attack. Therefore, aiming at the tuning and optimization of instantaneous attitude control parameters of high-speed loads, an improved particle swarm optimization algorithm is designed, which designs the adaptive adjustment method of inertial weights, so that different particles produce different inertia weights according to their own search effects, and the idea of genetic algorithm is introduced to expand the search range of particle swarms, so as to improve the optimization ability of particle swarm optimization algorithms, and obtain a high optimization success rate of instantaneous attitude control of high-speed loads under harsh initial conditions.
Key words: particle swarm algorithm     genetic algorithms     inertia weights
0 引　言

1 高速载荷瞬时姿态控制参数优化

 图 1 高速载荷姿态控制流程 Fig. 1 High-speed load attitude control process

 $U(s) = \left(K_P^\alpha + K_D^\alpha \dfrac{1}{{1 + K_N^\alpha \dfrac{1}{s}}}\right)(R(s) - R\_tar) 。$ (1)

2 粒子群优化算法的改进 2.1 基本粒子群优化算法

 $\left\{ {\begin{gathered} \begin{gathered} {v_{ij}}(t + 1) = \omega {v_{ij}}(t) + {c_1}{r_1}\left[ {pbes{t_{ij}} - {x_{ij}}(t)} \right] + \\ {c_2}{r_2}\left[ {gbes{t_j} - {x_{ij}}(t)} \right] ，\\ \end{gathered} \\ {{x_{ij}}(t + 1) = {x_{ij}}(t) + {v_{ij}}(t)} 。\qquad\qquad\qquad\ \ \end{gathered}} \right.$ (2)

①粒子群具有较大惯性权重。若种群中的粒子发现了某些较好的位置，但因粒子具有较大惯性权重，粒子的迭代搜索步长较大，使粒子无法在可能的最优解局部位置进行细致搜索，从而错过最优解。

②粒子群具有较小的惯性权重。种群中的粒子迭代步长较小时，会导致粒子只能在当前最优解的局部范围内搜索。但如果前期粒子群没有对全局进行较好的搜索，粒子群会过早陷入局部极值，从而造成当前的最优解与全局最优解有较大差距。

 \left\{\begin{aligned} &{{\omega _j}(i) = {\omega _{\min }} + ({\omega _{\max }} - {\omega _{\min }}) \times \displaystyle\frac{1}{{1 + 0.5 \cdot \exp [ - h({f_j}(i))]}}}，\\ &{h({f_j}(i)) = {{\tanh }^{ - 1}}[\frac{2}{{b - a}}({f_j}(i) - a) - 1]} 。\end{aligned} \right. (4)

 $\left\{ {\begin{array}{*{20}{c}} {a = f{{(i)}_{\min }}} ，\\ {b = f{{(i)}_{\max }}} 。\end{array}} \right.$ (5)

2）引入遗传算法

①变异操作

 ${\rm{mut}}\_{\mathrm{step}} = rand\times{(1 - i/\max \_{\mathrm{iteration}})^2}。$ (6)

 $populatio{n_{ij}} = populatio{n_{ij}}\times(1 \pm {\mathrm{mut}}\_{\mathrm{step}})。$ (7)

②交叉操作

③选择操作

 $({f_{\max }} - {f_{\min }})\times0.8 + {f_{\min }} 。$ (8)

 $f = \int_{0.5}^{1.5} {t(\arccos (\cos \left( \alpha \right)\cos \left( \beta \right)))} {\mathrm{d}t} 。$ (9)

2.3 基于改进粒子群算法的控制参数优化过程

 图 2 基于改进粒子群优化算法的控制参数优化流程 Fig. 2 Control parameters optimization process based on improved particle swarm optimization algorithm
3 仿真结果与分析

 图 3 遗传算法仿真结果 Fig. 3 Simulation results of genetic algorithm

 图 4 基本粒子群算法仿真结果 Fig. 4 Simulation results of elementary particle swarm algorithm

 图 5 改进粒子群算法仿真结果 Fig. 5 Simulation results of improved particle swarm algorithm

4 结　语

 [1] 智慧. 群智能优化算法的研究及应用[D]. 西安:西安电子科技大学, 2020. [2] 高岳林, 杨钦文, 王晓峰, 等. 新型群体智能优化算法综述[J]. 郑州大学学报（工学版）, 2022, 43(3): 21-30. [3] 安家乐, 刘晓楠, 何明, 等. 量子群智能优化算法综述[J]. 计算机工程与应用, 2022, 58(7): 31-42. DOI:10.3778/j.issn.1002-8331.2110-0141 [4] 张九龙, 王晓峰, 芦磊, 等. 若干新型智能优化算法对比分析研究[J]. 计算机科学与探索, 2022, 16(1): 88-105. [5] 韩礼红. 影响力最大化智能优化算法及应用研究[D]. 兰州:兰州大学, 2021. [6] HOLLAND J H. Adaptation in natural and artificial systems: an introductory analysis with applications to biology, control, and artificial intelligence[M]. 2nd ed. Cambridge: MIT Press, 1992. [7] KENNEDY J, EBERHART R. Particle swarm optimization[C]// IEEE International Conference on Neural Network. New Jersey, USA: IEEE, 1995, (4): 1942−1948. [8] 杨博雯. 粒子群优化算法的惯性权重研究[D]. 锦州:渤海大学, 2021. [9] 李会荣, 彭娇. 带有非线性惯性权重和柯西变异的粒子群优化算法[J]. 计算机与数字工程, 2021, 49(7): 1325-1329+1362. [10] 王生亮, 刘根友. 一种非线性动态自适应惯性权重PSO算法[J]. 计算机仿真, 2021, 38(4): 249-253+451.