Dynamic area coverage is widely used in military and civil fields. Improving efficiency an important research direction for multi-agent dynamic coverage. In this paper, we focus on the non-optimal problem of free algorithms. We propose a distributed algorithm based reinforcement learning <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\gamma $ </tex-math></inline-formula> -information map. The map can...

摘要: 针对传统卷积神经网络(Convolutional neural network,CNN)应用于焊缝缺陷识别时,池化模型适应差及特征选择能力低、以及易导致过拟合问题,提出了一种基于改进卷积神经网络(Improved pooling model and feature selection of CNN,IPFCNN)的焊缝缺陷识别方法。结合焊缝缺陷图像本身的特点,对传统平均池化模型做出改进,提出一种综合考虑池化域与其所在区域特征图分布的池化模型;为增强模型特征选择能力,提出将随机森林与卷积神经网络相结合的强化特征选择方法。以某汽轮机制造过程中焊缝缺陷识别案例对所提方法进行了验证和说明,结果表明提出的池化模型在处理不同特征分布的池化域时具有动态自适应性,并通过提高特征选择能力,使得所提方法比传统CNN方法具有更高的缺陷识别率。

The charging issue in Wireless Rechargeable Sensor Networks (WRSNs) is a popular research problem. With the help of wireless energy transfer technology, electrical can be from Charging Equipment (WCE) to sensor nodes, providing new paradigm prolong network lifetime. Existing usually takes periodical and deterministic approach, but ignore limited WCE influences non-deterministic factors such as topological changes node failures, making them unsuitable for real networks. In this study, we aim...

In recent years, with the development and wide application of unmanned aerial vehicle (UAV) technology, key technology network has become a new research focus. Compared traditional networks, UAV networks have characteristics dynamic topology limited energy. These brought challenges to design implementation communication systems. We need redesign organizational structure reduce power consumption, extend operation time, improve stability. this paper, we studied clustering designed strategy....

Abstract A study of the micromechanical behavior to predict overall response short fiber-reinforced composites under cyclic mechanical and thermal loading is presented. The instantaneous average over a “representative volume” material considered. influence fiber’s aspect ratio, volume fraction, spatial orientation has been investigated. linear combined hardening model used describe effects in case metal matrix. numerical procedure conditions. results have compared three different models...

To tackle challenges such as interference and poor accuracy of indoor positioning systems, a novel scheme based on ultra-wide bandwidth (UWB) technology is proposed. First, we illustrate distance measuring method between two UWB devices. Then, Taylor series expansion algorithm developed to detect coordinates the mobile node using location anchor nodes them. Simulation results show that observation error under our strategy within 15 cm, which superior existing algorithms. The final...

In this paper a strategy for increasing data persistency in wireless sensor networks has been proposed. Due to usually hazardous environment and limited resources, sensors the network are vulnerable. Therefore best approach increase is distributed storage. We propose an Efficient Distributed Storage based on LT codes (EDSLT) which implements directional random walk mechanism packet transmission, ID-based estimation method used estimate global information at phase, process of XORing packets...

This work mainly discusses the application of an improved fuzzy ARTMAP network based on logic in wetland classification. The main improvement is represented two phases: first category competition phase, reasoning replaces simply similarity criterion; second studying (resonance) not only winning neuron's weights but also membership functions are updated according to input characters. data source TM images Zhalong nature reserve north China, and both spectrum spatial characters extracted from...

The purpose of this paper is to apply a neural network classify aerial photographs. An adaptive backpropagation algorithm introduced build four-layer as the practical model photographs based on pixel-pixel. example classification lake, forest and land, comparing with maximum likelihood through analysis process results. in succeeded building complex an photograph it solved problem great storage data. back-propagation adopted speeded up learning rate, general error decreased smaller degree,...

Hierarchical autonomous system (AS) solves the problem of controlling entire Space Information Network (SIN) for whose high component complexity and dynamic. Aiming at that storage resource AS network's node severely limited, we propose an efficient Distributed Date Storage based on Fountain codes (DDSF) strategy. DDSF implements directional random walk mechanism packet transmission, ID-based estimation method is used to estimate global information during pre-coding phase, process XORing...

The wireless sensor network (WSN) is the development trend of technology sensor. This paper, based on nRF905 module, introduces a temperature gathering and transmitting system. From RF modules master control hardware platform has been designed, beyond that, paper software design. test show that system stable, data are reliable.

Focused on revealing variations of dynamic characteristics an auto-leveling system used in mobile elevating work platforms (MEWP) at large height, a unified mathematical model composited series sub-models was built for the leveling systems. An effect including fast response and relative stability from varying-parameters investigated. Results displayed displacement-angle factor changing with work-conditions tended to cause fluctuations stability, while enlarging amplifier gain would...

This paper focuses on stability analysis and stabilization of a switched system under average dwell time criteria in the continuous-time domain. The matrix polynomial approach is applied to construct continuous Lyapunov function design more effective controller, which can not only improve performance but also lessen conservativeness. problem studied with square matricial representation for first time. new method applicable than previous work switching time-varying controller gains. In...