摘要：Crickets, similar to some other insects, have highly sensitive filiform hairs on their cerci that can detect miniscule changes in airflow. This study imitates the perception mechanism of these filiform sensory hairs of crickets by designing and fabricating a Multi-electrode Metal Core Piezoelectric Fiber （MMPF）-based airflow sensor. Four longitudinal conductive sheets were coated symmetrically on their surfaces with Metal-core Piezoceramic Fibers （MPF）. The four fan-shaped piezoelectric ceramics with surface electrode covers were polarized. After successful polarization, the cantilevered MMPF could be used as an airflow sensor. The four electrodes on the surface were symmetrically divided into two groups. Therefore, two signals can be produced by a single fiber sensor. The theoretical model of an MMPF airflow sensor has been established. The model indicates that the ratio of the two signals is equivalent to the tangent of the airflow angle. Furthermore, the sum of the squares of the two signals is not dependent on the angle, but reflects the velocity of the airflow. Therefore, a single MMPF can be used to measure both the direction and amplitude for a given airflow. The theoretical model has been confirmed via experimental measurements.

摘要：在分析了汽车涡轮增压器进气管结构的基础上,根据三维吸入式吹塑模具结构的设计特点,基于Pro/E进行了汽车涡轮增压器进气管的吹塑模具设计。重点对分型面的选择、夹紧装置、吸口装置、吹针及冷却系统等关键部位进行了设计。实际生产表明,该模具在三维吸入式吹塑机上生产运行状态良好,生产的进气管质量能够达到设计要求,模具的正常寿命大于30万模次,具有良好的经济效益。

摘要：模仿昆虫毛发感受器的结构和功能,设计制备了一种能测量外部振动的表面四电极含金属芯压电纤维(FMPF)冲击振动传感器。采用干压成型法制备了含铂金芯的压电陶瓷纤维胚体,经过高温烧结、涂镀表面电极、高温极化后,制成FMPF传感器。基于压电方程和振动理论,建立了悬臂梁结构的FMPF冲击振动传感器的理论模型,分析了传感信号与冲击振动角度和幅值的关系。把FMPF固定在基体上,搭建了实验系统,测试了FMPF对冲击振动的响应,验证了理论模型。结果表明,FMPF的传感信号和冲击振动的幅值成线性关系,和方向成"8"字形关系。得出了FMPF传感器能够测量冲击振动幅值和方向的结论。

摘要：目的胸外按压及电除颤是救治心脏骤停患者的两种常用手段,对患者进行电除颤前需分析其心电图(electrocardiogram,ECG),而胸外按压会干扰患者的心电分析。本研究设计了一种用于心肺复苏的心电分析算法以及配套心电采集模块,以解决因胸外按压伪迹导致的心电误识别为室颤的问题。方法该心电采集模块采用若干模拟数字电路搭建而成。自动分析程序采用经近似熵算法改进的卷积神经网络(convolutional neural network,CNN)模型进行心电分类。采用MIT-BIH心律失常数据库进行验证,并在实机测试中通过在心电采集模块中耦合同频段干扰进一步验证。结果相比无近似熵改进版本的分类程序,使用该CNN模型分析程序可将含同频段伪迹的正常心电识别准确率由CNN的86.3%提升至97.78%;在存在同频段伪迹的实机测试中,使用该分析程序可将心电识别准确率由20%提升至96%。测试结果表明,当存在非同频段伪迹干扰时,该设计模块分类准确;当存在同频段伪迹的情况下,对正常心电信号的心电分类准确率依然高达90%以上。结论基于熵和卷积神经网络模块的抗按压伪迹干扰心电自动识别性高、抗干扰能力强,后续可推荐其用于自动胸外按压电除颤一体机等高同频段伪迹场合的心电采集与自动诊断,以减少误判。