地面运动装甲目标声震信号精细化特征提取与智能识别技术研究

项目名称： 地面运动装甲目标声震信号精细化特征提取与智能识别技术研究

项目编号： No.61501468

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 无线电电子学、电信技术

项目作者： 丁凯

作者单位： 中国人民解放军63983部队

项目金额： 22万元

中文摘要： 现代战争的战场环境日趋恶劣，多样化因素的综合影响增大了广域值守弹药系统对地面运动装甲目标精确识别的技术难度。传统的探测手段、识别准确率和效率已不能满足实际需求。为显著提高广域值守弹药系统对目标识别的准确度和可靠性，增强系统的抗干扰能力和环境适应能力，在地面运动装甲目标辐射声信号基础特性及地面震动波传播及响应特性分析的基础上，系统开展典型运动装甲目标声、震测试信号预处理技术研究，建立基于非线性特征、二代小波族特征及数学形态特征的典型地面运动装甲目标声、震测试信号的精细化描述模型，并在此基础上考虑到战场环境下的复杂干扰因素，构建基于声震数据融合技术和自适应神经模糊推理算法的目标智能识别系统。研究成果可直接运用于包括智能地雷在内的广域值守弹药系统中，提高我军智能武器的目标识别率和可靠性，同时对其它行业内复杂环境下多传感器目标识别技术的提高也具有重要意义和推广应用价值。

中文关键词： 装甲目标；声信号；震动信号；特征提取；智能识别

英文摘要： In modern warfare, the battlefield environment is increasingly complicated. The combined effects of diverse factors increases the technical difficulty of accurate identification of moving armored vehicles by wide-area guard ammunition(WAGA) systems. Traditional detection means, recognition accuracy and efficiency cannot meet the actual demand any more. In order to significantly improve the accuracy and reliability of the target identification system of WAGA, enhance the anti-jamming ability and adaptability to the environment of target identification, on the basis of analyzing the basic characteristics of acoustic signals and the response/propagation characters of vibration signals( radiated by armored vehicles), the preprocessing technology research of the acoustic and vibration signals of armored vehicles would be carried out systematically; A refined descriptive model on nonlinear feature, 2nd generation wavelets feature and mathematical morphology feature of the signals would be established. Considering the diverse and complicated interference factors in battlefield environment, the intelligent identification system for armored vehicles would be built based on data fusion technology of acoustic and vibration signals and adaptive neural fuzzy inference (ANFI) algorithm. The research results can be applied into the WAGA directly, including smart landmines, which will improve the recognition rate and reliability of our smart ammunitions. Meanwhile, the research achievements could promote the technology of multi-sensor target recognition within a complex environment in other industries, which has an important significance and application value.

英文关键词： armored vehicle;acoustic signal;vibrative signal;feature extraction; intelligent identification