基于TPAM-DPN网络的雷达辐射源识别方法

针对传统依靠人工经验设计并提取雷达辐射源特征的繁琐和区分度不够的问题,提出了一种改进的双路网络(DPN)自动提取特征并识别的方法。首先将一维的雷达时域信号变换到二维时频域,然后直接输入双路网络进行识别,即将雷达辐射源的识别转化为图像的识别,有效缓解了上述问题。同时,针对双路网络层次过深带来的特征流失问题,提出用于对雷达辐射源特征图校准重采样的轻量级模块——三流注意力模块(TPAM),并嵌入双路网络构成三流注意力双路网络(TPAM-DPN)对雷达辐射源进行识别。对6种常见的雷达信号进行了仿真实验,证明了所提方法提取的特征更有利于提高雷达辐射源识别率,且时效性更好。

一种基于变化检测技术的SAR图像舰船目标鉴别方法

该文引入变化检测思想,利用SAR图像中海杂波和目标之间的灰度差异,通过对潜在舰船目标切片的目标像素和背景像素进行分离,计算目标像素聚集度(TPAM)特征,实现对高亮像素在图像切片中聚集程度的定量评估,从而鉴别目标切片中是否包含有舰船目标,有效去除杂波虚警。首先,基于感兴趣区域(ROI)切片中心为目标像素及四周为海杂波的合理假设,构建似然比变化检测量获取差异图像;然后,利用KSW熵阈值选择方法实现差异图像中目标像素和海杂波像素的自动分离,生成二值图像;最后,利用切片中心像素为种子点,对二值图像进行区域生长,计算目标像素聚集度特征,并判断目标切片是否包含舰船目标。基于RADARSAT-1 SAR实测数据的实验结果表明,该文方法得到的目标像素聚集度特征计算简单、稳健性好、可区分度高,具有良好的鉴别性能,能够去除大部分海杂波干扰产生的虚警,有效地降低目标检测虚警率。

GIT变频器的设计与研究

随着理念与科技的进步,"绿色、智能、精致"是当今研发新型变频器的主要目标。因此,对GIT(Green Intelligent Type:"绿色智能型")变频器的研发更显得意义非凡。此文中运用现代电气电子、智能控制、变频调速、计算机、通信等理论与技术,对此GIT变频器进行了精心设计与研究,并着重论述了该GIT变频器设计中智能理论与技术的运用方法、要领和技巧,且撰写成此文,与同仁交流,以期抛砖引玉。

中学生考试焦虑研究

目的 作者运用自己设计的缓解中学生考试焦虑的心理调节训练方法 (Training method self-created of apsychology adjustmentto the middle school students简称 TPAMS训练法 )对中学生进行辅导和训练 ,探讨 TPAMS训练法是否能够有效地缓解中学生的考试焦虑。方法 随机选择泸州市某中学高三年级文、理科各 1个班共 2 64名学生 ,其中男生 1 1 8名 ,女生 1 48名作为研究对象 ,对研究对象在 TPAMS训练法训练前及训练后分别用考试焦虑测验量表 (TAT)进行测验并进行统计分析。结果 经过 TPAMS训练法训练后的男女学生考试焦虑水明显降低 ,有统计学意义。结论 TPAMS训练法对于缓解中学生的考试焦虑水平具有良好效果。

基于单片机的电动机智能软起动控制系统研究与设计

电动机智能软起动控制系统的应用范围较广,传统的电动机控制系统存在起动电流大、时间长等缺陷,亟待优化改进。现针对TPAM研究了基于单片机控制的软起动技术,采用晶闸管调压方式,通过改变晶闸管导通角来实现对TPAM定子两端电压的调节。

Relative position fitting based control strategy for thruster point assembly mechanism of electric propulsion

Purpose-Thruster point assembly mechanism(TPAM)of the electric propulsion system allows to adjust the thrust vector,sothat the thrust vectorisdirectedtothesatellitecenterof gravity(COG)during the satellite on-orbit working period.In this way the impact of disturbance torque caused by deviation of the thrust vector from the satellite COG during thruster ignition can be decreased.Therefore,the control accuracy of satellite is influenced directly by the control accuracy of TPAM.On the other hand,the on-orbit application of TPAM is restricted to the on-orbit computer resource.Therefore,the purpose of this paper is to design a control strategy for TPAM,and the strategy should not only be able to control the TPAM precisely but also be easily implemented by the on-board computer.Design/methodology/approach-First,the structure and work principle of TPAM are discussed,and the mathematical model based on D-H coordinate system is built for it.Then the fitting methods are utilized to design the control strategy of TPAM.Absolute position fitting-based control strategy and relative position fitting-based control strategy are designed,and the least squares algorithm is introduced for parameter selection.Findings-Simulations and tests are provided for the TPAM.Compared with the state-of-the-art PD controller,the proposed control strategy shows smaller overshoot and more simple realization.The experiment results are matched with the simulation results and both the experiment and simulation results show the validity of the proposed control strategies.Practical implications-The designed control strategies can be used for the TPAM of some satellite’s electric propulsion system.Originality/value-The mathematical model of the TPAM based on D-H coordinate system is given.The absolute position fitting-based control strategy and relative position fitting-based control strategy are proposed.Compared with existing methods,the two control strategies have more simple structure and smaller amount of computations.Furthermore,the relative position fitting-based control strategy achieves high precision with simple structure.