基于语音人工智能的“黑广播”监测方法研究

"黑广播"会扰乱正常信号,破坏空中电波秩序,对社会危害极大,因此,对"黑广播"进行监测与打击是无线电管理的重要工作之一。根据广播语音内容甄别"黑广播"是最可靠的识别方式,传统的"黑广播"识别依靠人工听取广播内容,成本高、效率低,基于自动语音识别技术的方式大多依托在线网络语音服务,难以应对离线实时监测场景。本文提出一套基于语音人工智能的"黑广播"监测识别技术,提取语音声学矩阵信息并利用神经网络进行识别,可在离线情况下实现"黑广播"的实时侦测。     

基于小波变换和概率神经网络的焊接接头类型识别

为了实现机器人自动焊接过程中快速、精确地提取焊缝特征信息,提出了一种基于小波变换和概率神经网络的焊接接头类型识别方法。先采用小波变换对由激光视觉传感器采集的焊接接头图像进行降噪和增强,对重构后的图像进行二值化,然后提取图像的特征信息,组成图像特征向量,最后构建概率神经网络分类器并进行测试。结合视觉传感器中激光器与摄像机的位置关系,最终识别出4种焊接接头。实验结果表明,所提出的方法特征提取简单,识别率高,并具有较好的实时性。     

Crystallization of Poly(ethylene adipate) within γ-Phase Poly(vinylidene fluoride) Matrix

The effects of PEA on the γ-phase PVDF crystal structure and the crystallization of PEA within the pre-existing γ-phase PVDF spherulites have been investigated by optical microscopy(OM), infrared spectroscopy(IR) and scanning electron microscopy(SEM). The results demonstrate that the γ-phase PVDF spherulites consist of the lamellae exhibiting a highly curved scroll-like morphology and develop preferentially in PEA-rich blend. With increasing PEA concentration, the scroll diameter increases and the scrolls are better separated from each other. PEA crystallizes first in the interspherulitic region and transcrystalline layer develops. Subsequently, the transcrystalline layer of PEA continues to grow within the γ-phase PVDF spherulites, e.g., in the region between the scrolls, until impinging on other PEA transcrystalline layers or spherulites. The crystallization kinetics results indicate that the growth rate of PEA crystals in the intraspherulitic region of γ-phase PVDF shows a positive correlation with content of PEA, but a negative one with the crystallization temperature of γ-phase PVDF.     

Mass and isotopic concentrations of water-insoluble refractory carbon in total suspended particulates at Mt.Waliguan Observatory(China)

Mass concentration and isotopic values δ¹³C and ¹⁴C are presented for the water-insoluble refractory carbon（WIRC） component of total suspended particulates（TSP）,collected weekly during 2003,as well as from October 2005 to May 2006 at the WMO-GAW Mt.Waliguan（WLG） site.The overall average WIRC mass concentration was（1183±120）ng/m³（n = 79）,while seasonal averages were 2081 ±1707（spring）,454±205（summer）,650±411（autumn）,and 1019±703（winter） ng/m3.Seasonal variations in WIRC mass concentrations were consistent with black carbon measurements from an aethalometer,although WIRC concentrations were typically higher,especially in winter and spring.The δ¹³C PDB value（-25.3 ± 0.8）%.determined for WIRC suggests that its sources are C₃ biomass or fossil fuel combustion.No seasonal change in δ¹³C PDB was evident.The average percent Modern Carbon（pMC） for ¹⁴C in WIRC for winter and spring was（67.2 ± 7.7）%（n = 29）.Lower pMC values were associated with air masses transported from the area east of WLG,while higher pMC values were associated with air masses from the Tibetan Plateau,southwest of WLG.Elevated pMC values with abnormally high mass concentrations of TSP and WIRC were measured during a dust storm event.