Promising electromechanical energy converters based on new materials and coatings

The efficiency of new materials and the application of coatings for promising high-speed electromechanical energy converters has been estimated. A new basic model of miniturbogenerator intended for testing new materials and coatings has been developed. The experimental results obtained for pilot samples with different organic-silicate coatings show that the losses in the rotor decrease.     

Development and trends of applications of the new types of turbogenerators with vector system of field control

The results of the application of asynchronized turbogenerators with vector control and biaxial excitation in the Russian electric power system are presented. Since 2003, five air-cooled asynchronized turbogenerators with powers of 110, 160, and 320 W have been in operation at heat electric power stations. The main results of their testing and operation are given.     

Powers and speeds in electromechanical systems: Hybrid propulsion systems in cars

Mathematical expressions for determining the moments and speeds of the elements of planetary gearboxes are obtained. The possible modes of the operation of an internal combustion engine (ICE) combined with two electric machines are shown. The relation between powers of electric machines and the geometrical parameters of the planetary gear is achieved.     

四合一天线合路器的控制与实现

介绍了一种基于FPGA控制的四合一天线合路器,该合路器能够根据4部电台的不同频率选择正确的通信信道,并将其信息传送到一路天线上进行通信。文中介绍了该合路器的硬件结构,软件控制算法以及工程上的实现方法。测试表明,该天线合路器能够满足4部电台同时在一路天线上进行接收与发送的要求,具有广泛的实用价值。     

矿用组合开关控制执行机构误动作的预防及处理

煤矿井下使用的矿用隔爆兼本质安全型多功能真空组合开关（以下简称组合开关）,对主回路的控制大多数选用PLC控制继电器完成对主回路的保护。但由于继电器机械触点抖动或组合开关机械自身故障等原因,PLC虽发出了动作指令,而执行机构容易发生误动作,从而降低了系统的可靠性。简要论述了组合开关的工作原理、执行机构误动作的分析预防和措施。     

基于Leap Motion的动态手势识别方法

伴随虚拟现实（Virtual Reality,VR）技术的发展,以及人们对人机交互性能和体验感的要求提高,手势识别作为影响虚拟现实中交互操作的重要技术之一,其精确度急需提升^[1].针对当前手势识别方法在一些动作类似的手势识别中表现欠佳的问题,提出了一种多特征动态手势识别方法.该方法首先使用体感控制器Leap Motion追踪动态手势获取数据,然后在特征提取过程中增加对位移向量角度和拐点判定计数的提取,接着进行动态手势隐马尔科夫模型（Hidden Markov Model,HMM）的训练,最后根据待测手势与模型的匹配率进行识别.从实验结果中得出,该多特征识别方法能够提升相似手势的识别率.