A narrow-linewidth continuous wave Ho：YAL03 laser with Fabry-Perot etalons

A narrow linewidth continuous wave Ho：YAP laser with two Fabry-Perot etalons pumped by a Tm：YLF laser is reported. The maximum output power reaches 8.3 W when the incident pump power is 15.8 W, with 52.5% optical-to- optical conversion efficiency and 62.6% slope efficiency. A stable laser output at 2118.1 nm is achieved, with a linewidth less than 0.4 nm （full width at half maximum）. The beam quality factor is M2- 1.25, measured by the traveling knife-edge method.     

A Continuous-Wave Ho：YA103 Laser with Output 8.5 W Pumped by a 1.91μm Laser at Room Temperature

A cw high efficient Ho：YAI03 laser pumped by 1.91 μm diode-pumped Tm：YLF laser at room temperature is realized. The maximum output power reaches 8.5 W when the incident pump power is 15.6 W. The slope efficiency is 63.7%, and the Tm：YLF to Ho：YAP optical conversion efficiency is 54.5%. The laser wavelength is 2118.3nm when the transmission of output coupler is 30%. The beam quality factor is M2 -1.39 measured by the traveling knife-edge method.     

A narrow-linewidth continuous wave Ho:YALO3 laser with Fabry–Perot etalons

A narrow linewidth continuous wave Ho:YAP laser with two Fabry–Perot etalons pumped by a Tm:YLF laser is reported. The maximum output power reaches 8.3 W when the incident pump power is 15.8 W, with 52.5% optical-tooptical conversion efficiency and 62.6% slope efficiency. A stable laser output at 2118.1 nm is achieved, with a linewidth less than 0.4 nm(full width at half maximum). The beam quality factor is M2～ 1.25, measured by the traveling knife-edge method.     

基于超声波反射系数相移的油膜厚度测量方法研究

超声波液相厚度测量技术已经在柴油机气缸油膜测量领域中得以验证,并得到了一定的应用. 然而,当油膜厚度变化尺度较大时,常规谐振模型与弹簧模型之间存在较大的测量盲区,使超声波测量技术的应用受到较大程度的限制. 基于超声波反射系数相移的油膜厚度测量方法可突破两种模型在大尺度油膜厚度测量方面的局限性,拓展超声波油膜厚度测量范围. 通过开展超声波反射系数相移机理研究,首次建立油膜两侧介质声阻抗相同时的直接求解方法;搭建油膜厚度超声波反射系数相移测量系统,利用单一中心频率对10~90 μm的油膜厚度展开测量,测量厚度与设置厚度具有很好的一致性,从理论和实验角度证明基于超声波反射系数相移的油膜厚度测量方法在大尺度油膜厚度测量方面的适用性.      

Judd-Oflet analysis of spectrum and laser performance of Ho：YAP crystal end-pumped by 1.91μm Tm：YLF laser

The Ho:YAP crystal is grown by the Czochralski technique.The room temperature polarized absorption spectra of Ho:YAP crystal was measured on a c cut sample with 1 at% holmium.According to the obtained Judd-Ofelt intensity parameters Ω2 = 1.42 × 10-20 cm2,Ω4 = 2.92 × 10-20 cm2,and Ω6 = 1.71 × 10-20 cm2,this paper calculated the fluorescence lifetime to be 6 ms for 5I7 →5 I8 transition,and the integrated emission cross section to be 2.24×10-18 cm2.It investigates the room temperature Ho:YAP laser end pumped by a 1.91 μm Tm:YLF laser.The maximum output power was 4.1 W when the incident 1.91 μm pump power was 14.4 W.The slope efficiency is 40.8%,corresponding to an optical to optical conversion efficiency of 28.4%.The Ho:YAP output wavelength was centred at 2118 nm with full width at half maximum of about 0.8 nm.     

基于可调谐激光吸收光谱技术的柴油机排放NO浓度瞬态测量

NO_x是柴油机排放的主要污染物质,它对人体和环境的危害极大,其中NO含量又占NO_x总量的90％以上,因此合理准确的检测柴油机排放物中NO的含量对NO_x含量的检测尤为关键。现有的测试方法由于采用采样分析法,不能及时反映真实情况,光学检测方法作为一种新型的测试技术,在很多领域都有应用。结合光学检测方法对稳态和瞬态过程下的NO浓度用TLAS技术开展研究,通过模拟NO浓度的变化情况,验证这种方法在柴油机上应用的可行性,选择一台中心波长为5 263 nm的带间级联激光器用于检测,为获得不同浓度的混合气体设计一套配气混合系统,用高纯度的ＮＯ作为被测气体,N₂作为背景气体,配比不同浓度的测试样气,在稳态测量时,分别配比浓度为500,1 000,2 000 ppm的样气进行试验,得到相对误差小于1%,最大绝对误差为11.5 ppm。在瞬态测量中,分别对浓度增加与减少过程进行试验,实时测量气室内的浓度变化情况,在1 ms时间分辨率的情况下可实现5.2%的最大相对误差。通过设计的系统,可以实现NO浓度的瞬态测量,能将此方法应用到柴油机上,这种方法有利于研究柴油机的排放物含量。     

利用可调谐激光吸收光谱技术的柴油机排放温度测试研究

针对D4114B型柴油机排放尾气中的CO2气体开展测量研究,计算分析气体的体积分数以及温度。文中以可调谐半导体激光器吸收光谱(Tunable Diode Laser Absorption Spectroscopy,TDLAS)技术原理为基础,利用MATLAB中SIMULINK库中的各个模块,模拟尾气测量的实际过程。仿真结果显示,在模拟柴油机排放环境下,待测量气体CO2的温度仿真相对误差为0.03%。利用船用D4114B型柴油机进行验证实验,在其排气管上增添可视化窗口并安装相应测试系统,利用以半导体为工作介质的可调谐激光器作为激光光源,开展尾气排放中CO2气体温度的在线测试研究,测试相对误差小于4%。由上述研究结果可知,本文中利用SIMULINK搭建的模型所测得的温度值与实际柴油机尾气排放过程中的温度相差较小,因此,其仿真结果能够对柴油机排气测温提供一定的参考。     

Exploring Explicit Coarse-Grained Structure in Artificial Neural Networks

We propose to employ a hierarchical coarse-grained structure in artificial neural networks explicitly to improve the interpretability without degrading performance. The idea has been applied in two situations. One is a neural network called Taylor Net, which aims to approximate the general mapping from input data to output result in terms of Taylor series directly, without resorting to any magic nonlinear activations. The other is a new setup for data distillation, which can perform multi-level ...