Task assignment in ground-to-air confrontation based on multiagent deep reinforcement learning

The scale of ground-to-air confrontation task assignments is large and needs to deal with many concurrent task assignments and random events.Aiming at the problems where existing task assignment methods are applied to ground-to-air confrontation,there is low efficiency in dealing with complex tasks,and there are interactive conflicts in multiagent systems.This study proposes a multiagent architecture based on a one-general agent with multiple narrow agents(OGMN)to reduce task assignment conflicts.Considering the slow speed of traditional dynamic task assignment algorithms,this paper proposes the proximal policy optimization for task assignment of general and narrow agents(PPOTAGNA)algorithm.The algorithm based on the idea of the optimal assignment strategy algorithm and combined with the training framework of deep reinforcement learning(DRL)adds a multihead attention mechanism and a stage reward mechanism to the bilateral band clipping PPO algorithm to solve the problem of low training efficiency.Finally,simulation experiments are carried out in the digital battlefield.The multiagent architecture based on OGMN combined with the PPO-TAGNA algorithm can obtain higher rewards faster and has a higher win ratio.By analyzing agent behavior,the efficiency,superiority and rationality of resource utilization of this method are verified.

Research on virtual entity decision model for LVC tactical confrontation of army units

According to the requirements of the live-virtual-constructive(LVC)tactical confrontation(TC)on the virtual entity(VE)decision model of graded combat capability,diversified actions,real-time decision-making,and generalization for the enemy,the confrontation process is modeled as a zero-sum stochastic game(ZSG).By introducing the theory of dynamic relative power potential field,the problem of reward sparsity in the model can be solved.By reward shaping,the problem of credit assignment between agents can be solved.Based on the idea of meta-learning,an extensible multi-agent deep reinforcement learning(EMADRL)framework and solving method is proposed to improve the effectiveness and efficiency of model solving.Experiments show that the model meets the requirements well and the algorithm learning efficiency is high.

From Innocence to Suffering to Awareness:Nada Confrontation in Francis Macomber

The Short Happy Life of Francis Macomber is a quintessential Hemingway tale of one man's attempt to overcome an in ternal struggle by mastering the external world. Francis Macomber discovers his own bravery and strength when he ignores his self-consciousness and relies on instinct. This essay will examine Hemingway's code and how it confronts nada thus analyze Macomber's change from innocent to suffering to aware.

Tensor-Centric Warfare V: Topology of Systems Confrontation

In this paper, as a new contribution to the tensor-centric warfare (TCW) series [1] [2] [3] [4], we extend the kinetic TCW-framework to include non-kinetic effects, by addressing a general systems confrontation [5], which is waged not only in the traditional physical Air-Land-Sea domains, but also simultaneously across multiple non-physical domains, including cyberspace and social networks. Upon this basis, this paper attempts to address a more general analytical scenario using rigorous topological methods to introduce a two-level topological representation of modern armed conflict;in doing so, it extends from the traditional red-blue model of conflict to a red-blue-green model, where green represents various neutral elements as active factions;indeed, green can effectively decide the outcomes from red-blue conflict. System confrontations at various stages of the scenario will be defined by the non-equilibrium phase transitions which are superficially characterized by sudden entropy growth. These will be shown to have the underlying topology changes of the systems-battlespace. The two-level topological analysis of the systems-battlespace is utilized to address the question of topology changes in the combined battlespace. Once an intuitive analysis of the combined battlespace topology is performed, a rigorous topological analysis follows using (co)homological invariants of the combined systems-battlespace manifold.

Traps Confronting O2O Entrepreneurs

O2O industry is surging in 2015,while 020 companies face fiercer competition.Though those startups got several rounds of financing,they still need to solve different problems one by one and move on.1Vicious competitionHunter poaching bears the brunt to become the most severe challenge for O2O entrepreneurs.

Help Those Missing from the Table——Confronting Gender Issues in New Looks

The widespread application of new technologies,while empowering women with new opportunities,might also put them at disadvantage.For example,in comparison with males,the application of AI might be more likely to cost them their jobs.Meanwhile,women are missing out the opportunity to participate in the policy-making process–they are absent from the table.If no change is made in the current policies,we will miss the goal to achieve gender equality,the fifth of the 17 Sustainable Development Goals set for UN’s 2030 Agenda for Sustainable Development,warned elite women scientists with the Organization for Women in Science for the Developing World(OWSD).What shall be done now?How can we make a difference?They are in action to help.

无人集群博弈对抗系统仿真验证及决策关键技术综述

无人集群博弈对抗是一种新兴的作战样式,在智能化战争扮演着至关重要的作用,其核心是自主生成博弈对抗决策序列,为集群“赋能”。分析了无人集群博弈对抗系统仿真验证的进展;从基于专家系统和博弈论的技术、基于群体智能和优化理论的技术,以及基于神经网络和强化学习的技术三个方面论述了自主决策关键技术,以及课题组在自主决策上开展的相关工作;提出了无人集群博弈对抗的发展方向。

糖尿病患者自我厌恶现状及影响因素分析

目的了解糖尿病患者自我厌恶现状及其影响因素,为医护人员采取针对性干预改善糖尿病患者心理健康提供参考。方法采用一般情况调查表、自我厌恶量表、医学应对问卷和领悟社会支持量表对274例2型糖尿病患者进行调查。结果糖尿病患者自我厌恶得分(35.46±6.39)分,年龄、糖尿病相关并发症、回避应对、领悟社会支持总分是自我厌恶的主要影响因素(均P<0.05),可预测总变异的49.6%。结论糖尿病患者自我厌恶为中等偏高水平,护理人员应加强对糖尿病患者自我厌恶水平的评估和筛查,加强对年轻、有糖尿病相关并发症的患者的干预,帮助患者减少其回避行为,并增强其领悟社会支持水平,从而降低其自我厌恶水平。

边缘对抗结合层次门控卷积的人脸修复研究

针对目前人脸缺损图像修复存在边缘模糊和失真等问题,提出了一种两阶段层次门控卷积网络(Hierarchical gated convolutional network,HGCN),并将其与边缘对抗网络相结合用于人脸图像修复。首先,采用边缘对抗网络生成边缘图像。其次,将边缘图、掩模和缺损图像作为输入,训练HGCN的GAN模型以修复缺损人脸图像。HGCN网络采用门控卷积取代传统卷积,并引入了扩张卷积,网络的主体由粗修复模块和精修复模块组成。在粗修复模块中,编码器和解码器网络结构用于粗修复;在精细修复模块中,引入注意力机制来增强特征提取能力,进一步细化修复结果。实验使用Celeba-HQ数据集和NVIDIA不规则掩码数据集作为训练数据集,采用门控卷积网络和注意力机制网络作为实验对比模型,PSNR、 SSIM和MAE作为实验评估指标。实验结果表明,对于缺损区域小于20%的人脸图像,所提出的网络在上述三个指标上优于两种比较网络,而对于缺失区域大于20%的图像,所提出的网络与两种比较方法性能指标接近。在视觉效果方面,所提出的方法在细节上也优于两种对比网络。因而,所提出的网络可以明显提高图像修复效果,尤其是对图像细节的修复效果。

攻防对抗视角下的网络安全主动防御体系研究

为了解决纵深防御体系中防御能力固化、灵活性不足等问题,从攻防对抗视角出发,提出一种网络安全主动防御体系。首先,基于情报收集、监测预警、防御指挥和防御行动等要素的详细描述,设计了用于规划主动防御系统的能力组成框架;其次,选择网络杀伤链、ATT&CK知识库和OODA循环法则,设计了一个以攻助防的防御模型,并分析了引入博弈论进行防御算法设计的优势;最后,以实现为目标,构想了一个典型应用场景,以期为安全能力的体系化设计提供有益借鉴。

Data augmentation method for insulators based on Cycle-GAN

Data augmentation is an important task of using existing data to expand data sets.Using generative countermeasure network technology to realize data augmentation has the advantages of high-quality generated samples,simple training,and fewer restrictions on the number of generated samples.However,in the field of transmission line insulator images,the freely synthesized samples are prone to produce fuzzy backgrounds and disordered samples of the main insulator features.To solve the above problems,this paper uses the cycle generative adversarial network(Cycle-GAN)used for domain conversion in the generation countermeasure network as the initial framework and uses the self-attention mechanism and channel attention mechanism to assist the conversion to realize the mutual conversion of different insulator samples.The attention module with prior knowledge is used to build the generation countermeasure network,and the generative adversarial network(GAN)model with local controllable generation is built to realize the directional generation of insulator belt defect samples.The experimental results show that the samples obtained by this method are improved in a number of quality indicators,and the quality effect of the samples obtained is excellent,which has a reference value for the data expansion of insulator images.