复杂约束下的导弹作战任务规划模型和算法研究

导弹作战任务规划是一个涉及时间、资源、质量和其他关系约束的复杂问题。首先通过定义约束满足效用对基本约束满足模型进行了扩展，建立了导弹任务规划的约束满足优化模型。在此基础上，研究了任务规划模型求解的时间和效用传播算法，提出了基于综合效用的优化求解框架。该模型和求解框架易于解决具有多种约束因素的复杂问题，具有较好的通用性。通过定义软、硬约束效用，使得实际任务规划问题求解具有更好的灵活性。     

地铁火灾应急预案的制定

介绍地铁火灾发生时的几个主要的特点以及其巨大的危害性,引出了地铁火灾应急预案制定模式的重要性。通过对已有的资料进行研究、分析,同时结合实际,对地铁火灾应急预案制定时所包含的主要内容进行了科学的完善,得出了一个完整的预案制定模式。     

Operations planning revisited: theoretical and practical implications of methodology

Parts of NATO’s contemporary planning framework called the comprehensive operations planning directive (COPD), and parts of the operation-level planning process should be revised since they suffer from methodological inconsistency. This claim is defended by discussing contradicting methodological properties and heuristics applied when framing and managing a military problem in accordance with the COPD. The methodological inconsistency within the COPD; in other words, simultaneously applying contradictory methodological properties, implies one theoretical and three practical implications. The theoretical implication is summarised in a meta-theoretical framework and explained by discussing five methodological properties: non-linearity, emergence, independently changeable generalisations, invariance and boundaries. The three practical implications of methodology imply that methodology is guiding: the problem-frame, conceptual development and action. To improve military planners’ understanding and management of these four identified implications, NATO is recommended to develop a “handbook of methodology.” The purpose of such a handbook should be to emphasise the utility of methodology when planning military operations.     

三关节机器人的神经网络补偿自适应控制器设计

三关节机器人广泛用于工业生产、轮式或履带式排爆机器人,为了补偿由于机器人结构参数、作业环境干扰等不确定性因素造成的机器人动力学模型的不确定性,将机器人动力学模型分解为名义模型和误差模型两部分,其误差模型采用RBF神经网络进行补偿,得到其估计信息,神经网络的输出权值根据Lyapunov稳定性理论采用自适应算法进行调整。所设计的神经网络补偿自适应控制器解决了不确定性机器人动力学系统控制器设计的不确定性问题,同时,通过定义Lyapunov函数,证明了控制器能渐近、稳定地跟踪期望轨迹。机器人的3个关节在控制器的作用下,约在5 s时达到期望轨迹,神经网络约在5 s时逼近机器人动力学模型的误差模型,实验结果表明了机器人关节对期望轨迹具有良好的轨迹跟踪性能。     

Implementing defence policy: a benchmark-“lite”

Most countries put significant amounts of time and effort in writing and issuing high-level policy documents. These are supposed to guide subsequent national defence efforts. But do they? And how do countries even try to ensure that they do? This paper reports on a benchmarking effort of how a few “best of breed” small- to medium-sized defence organisations (Australia, Canada, and New Zealand) deal with these issues. We find that most countries fail to link goals to resources and pay limited attention to specific and rigorous ex-ante or post-hoc evaluation, even when compared to their own national government-wide provisions. We do, however, observe a (modest) trend towards putting more specific goals and metrics in these documents that can be – and in a few rare cases were – tracked. The paper identifies 42 concrete policy “nuggets” – both “do’s and don’ts” – that should be of interest to most defence policy planning/analysis communities. It ends with two recommendations that are in line with recent broader (non-defence) scholarship on the policy formulation-policy implementation gap: to put more rigorous emphasis on implementation (especially on achieving desired policy effects), but to do so increasingly in more experiential (“design”) ways, rather than in industrial-age bureaucratic ones (“PPBS”-systems).     

空中多飞行器飞行轨迹优化数值仿真

针对空中多飞行器在复杂环境中飞行轨迹的多目标最优问题，分析了多飞行器飞行过程中各种可视和不可视约束条件。基于在回避威胁区前提下燃料消耗最少、飞行时间最短的综合性能指标，采用“多方法组合”思路，提出了改进动态规划法和多点边值法组合算法，并进行了仿真验证，大量C＋＋数值飞行仿真结果表明该算法能够在考虑外界复杂环境和飞行器各种约束条件下快速规划出空中多飞行器的最优飞行轨迹，该组合算法具有一定的实用性和创新性。     

基于 ANP 的作战筹划能力模糊评价方法

基于网络层次分析（ANP）的模糊评价方法，通过对作战筹划能力进行分析评价，为提升作战筹划能力提供参考。首先运用 ANP 法，提出包含4个方面18个指标的评估指标体系，构建作战筹划能力评估模型，然后通过模糊综合评价法对作战筹划能力进行评价。该评估模型探索了筹划能力评价的方法，为作战筹划能力评价工作提供新思路。     

威胁联网下低空突防航路规划研究

分析了威胁联网下信息交流和资源共享对飞行航路规划的影响;针对威胁联网,制定相关的威胁体相互支援表,采用遗传算法进行低空突防航路规划.通过仿真计算证明,此方法规划出的飞行航路能有效提高威胁联网下的战斗机低空突防安全性,为航路规划提供了一种新思路.     

激光制导炸弹弹道拟合模型与算法

由于激光制导炸弹的弹道受控,其弹道复杂,采用传统炸弹弹道的拟合方法达不到精度要求。从抛物运动的角度提出了两种对激光制导炸弹弹道拟合的方法。第一种方法对一般的抛物运动标准解引入修正项实现对激光制导炸弹弹道的拟合。第二种方法对激光制导炸弹的攻击运动过程进行分析,然后,引入加速度修正项对激光制导炸弹弹道进行拟合。通过大量的仿真表明这两种方法的拟合精度均很高。相比较而言,第二种方法精度更高。