Pavement roughness is one of the most important factors for appraising highway construction. In this paper, we choose the laser triangulation to measure pavement roughness. The principle and configuration of laser triangulation are introduced. Based on this technology, the pavement roughness of a road surface is measured. The measurement results are given in this paper. The measurement range of this system is 50 μm. The measurement error of this technology is analyzed. This technology has an important significance to appraise the quality of highway after completion of the workload.

Self-mixing interferometry (SMI) based on nanometer fringes and polarization flipping is realized. The interferometer comprises a single-mode He-Ne laser and a high-amplitude reflectivity feedback mirror. The nanometer fringes are obtained by tilting the external feedback mirror. The fringe density is 35 times higher than that derived with conventional two-beam interference, and each fringe corresponds to a \lambda/70 displacement in external cavity length. Moreover, polarization flipping occurs when the external feedback mirror moves in the opposite direction. Such movement can be used to easily distinguish displacement direction. Experimental results show an optical resolution of displacement measurement of 9.04 nm with a range of 100 μm. The proposed SMI presents promising application prospects in precisely measuring displacement and calibrating other micro-displacement sensors because of its optical wavelength traceability.

The polarization state is modulated by tilting birefringence component placed in the feedback external cavity. The variation of the polarization state in one period of modulation is found to be similar to sine wave. The periods become increasingly smaller. The maximum of variation in one period decreases against the rotated angle. The experimental phenomenon is subjected to the change of optical path and secondary reflection. The phenomenon is analyzed theoretically based on geometrical optics and crystal optics. High-accuracy measurements of absolute and relative angles can be realized based on the experimental phenomenon. The angle resolution is 0.1 arcsec in theory.

提出了一种新颖的基于人工神经网络(ANN)和遗传算法(GA)的激光器参数全局优化方法,建立激光器输出功率的人工神经网络模型,来模拟激光器参数对输出功率的综合影响机理,进而以该模型作为目标函数,采用遗传算法对激光器参数进行全局优化。以平凹腔单横模氦氖激光器为例验证了该方法的可行性和有效性。对相同参数的激光器,人工神经网络模型的仿真数据与实验数据的均方根误差为0.0127 mW。应用该方法对其他参数全局优化后激光器预期输出功率比实验室已有的同等尺寸的激光器大,说明了该方法的有效性。