探测点序列的最优化与简化计算方法

在搜索状态建模和求解一阶搜索状态方程的特征迹线解的基础上,建立了对随机运动进行离散时间探测的发现概率最优控制模型,结合动态规划原理给出一种最优探测点序列的逼近算法,并给出了短时计算的算法简化形式。在满足一阶搜索状态方程的随机恒速目标条件以及有限指数探测函数条件下,将给出的算法及其简化形式应用到算例。算例表明,当随机恒速运动目标初始位置和速度均服从圆正态分布时,该算法及其简化形式均能够由任意给定的初始探测点序列优化收敛到满足精度要求的最优探测点序列。     

吊放声纳应召搜潜效能的建模与仿真

基于反潜直升机吊放声纳的探测原理及战术使用特点,在目标初始位置概略已知,速度和航向未知的条件下,分析了目标散布规律,建立了搜潜效能分析计算模型,给出了吊放声纳在扩展方形搜索方式下的应召搜潜概率；并采用蒙特卡洛法的基本思想对该计算模型进行了仿真验证,证明了该方法的有效性；最后,基于扩展方形搜索方式,研究了在单机和双机搜潜条件下搜索概率的不同结果,分析了潜艇速度、应召延迟时间、吊放次数等参数 对搜索概率的影响。     

基于冲突分类与消解的多智能体路径规划算法设计CSCD

多智能体路径规划应用广泛但求解困难。为更好地处理多智能体路径规划中的路径冲突问题,提高求解效率,将冲突进一步分类为相向顶点冲突和交叉顶点冲突,并提出了对应的消解方式。相向顶点冲突的消解方法采用提前添加约束的方式,避免在消解其冲突的过程中产生另一个可预见的冲突;交叉顶点冲突的消解方法采用寻找最佳等待时间的方式,在消解其冲突的同时消解其他存在的冲突。两种冲突消解方法均可减小约束树的规模,在一定程度上减少算法的计算量。并提出了基于冲突搜索算法的高层节点冲突搜索算法。实验结果表明,所提出的冲突分类及消解方式有效地减小了算法高层中约束树的规模,降低了算法计算量,并在智能体密集的环境下表现出更大的优势。     

未知动态环境下有人/无人机协同搜救研究

针对未知动态环境下搜救效率和智能化水平不高的问题，提出了一种基于变步长模型预测控制的有人机/无人机协同搜救方式。根据环境的动态变化性和探测的不确定性，建立了表示环境确定度和目标探测概率的动态搜救图模型，根据态势信息选择时间域步长，并根据有人机/无人机协同搜救任务特点，设计了搜救收益指标，并利用粒子群优化进行求解。仿真结果表明了本文方法的有效性和优越性。     

一种多目标变邻域模拟退火算法及成像星座任务规划方法

为提高成像星座观测任务的完成度和成像质量，提出一种多目标变邻域模拟退火算法(MOVNSA)。首先，分析了敏捷成像卫星观测任务约束和星上资源约束，构造了评价任务完成度和成像质量的目标函数，建立成像任务约束满足模型。然后通过设计多目标模拟退火算法中解的编码解码规则、变邻域搜索方法以及选拔淘汰机制，实现了观测任务成像时刻的滑动优化，兼顾提升了观测任务的完成度和成像质量。最后通过工程实例对多种算法进行测试和分析，结果表明该方法收敛速度快，优化效果好，可以在满足用户需求的前提下最大化任务的完成度。     

Experimental and numerical investigation of mechanical behavior of plain woven CFRP composites subjected to three-point bending

The mechanical behavior of plain woven Carbon Fiber-Reinforced Polymer (CFRP) composites under Three-Point Bending (TPB) is investigated via experimental and numerical approaches. Multiscale models, including microscale, mesoscale and macroscale models, have been developed to characterize the TPB strength and damages. Thereinto, Representative Volume Elements (RVEs) of the microscale and mesoscale structures are established to determine the effective properties of carbon-fiber yarn and CFRP composites, respectively. Aimed at accurately and efficiently predicting the TPB behavior, an Equivalent Cross-Ply Laminate (ECPL) cell is proposed to simplify the inherent woven architecture, and the effective properties of the subcell are computed using a local homogenization approach. The macroscale model of the TPB specimen is constructed by a topology structure of ECPL cells to predict the mechanical behavior. The TPB experiments have been performed to validate the multiscale models. Both the experimental and numerical results reveal that delamination mainly appears in the top and bottom interfaces of the CFRP laminates. And matrix cracking and delamination are identified as the significant damage modes during the TPB process. Finally, the quasi-static and dynamic behaviors of plain woven composites are discussed by comparing the results of Low-Velocity Impact (LVI) and TPB simulations.     

Local empirical model of ionospheric variability

The paper presents an analysis of the ionospheric variability as a function of local time, month, and geomagnetic activity level. The 2003–2020 dataset of peak electron densities (NmF2) from the Irkutsk DPS-4 Digisonde (52.3°N, 104.3°E) was converted into the dataset of the NmF2 disturbances (ΔNmF2) representing the relative (percentage) deviations of the NmF2 from the 27-day running median. The ΔNmF2 dataset was used to calculate root mean square values of ΔNmF2 (σNmF2) by 27-day running averaging. These σNmF2 values were considered as a measure of ionospheric variability. The σNmF2 as function of local time, day of year, and year was the input for building the local empirical model of ionospheric variability based on the linear regression of σNmF2 on the 27-day average daily Ap-index of geomagnetic activity. The paper demonstrates the diurnal-seasonal variations in σNmF2 under low geomagnetic activity (linear regression intercept) as well as the rate of increase/decrease in σNmF2 with increasing Ap (linear regression slope). The obtained diurnal, seasonal, and geomagnetic activity behavior of σNmF2 is compared with previous studies of ionospheric variability.