Near ultraviolet (λ ≈ 400 nm) femtosecond laser annealing (400 nm-FLA) in a scanning mode was employed to crystallize amorphous silicon (a-Si) films at room temperature. The average grain size of polycrystalline silicon annealed was studied as a function of the incident laser fluence and beam overlap or the number of laser shots irradiated. In general, the grain size can be enlarged by either increasing the beam overlap at a fixed laser fluence or increasing the laser fluence for a fixed number of laser shots. An apparent threshold for the onset of rapid enlargement of grain size was observed for processing at ∼90% overlap and fluences above 25 mJ/cm2. A maximum grain size of ∼280 nm was attained at a laser fluence of 30 mJ/cm2 and overlap of 93.75%, beyond which the grain size attained was smaller, and eventually, ablation was observed at an overlap of 97.5% and higher. These trends and observed surface morphology of annealed samples suggest that the crystallization mechanism is like sequential lateral solidification, similar to 800 nm-FLA and excimer laser annealing. Raman spectroscopic studies show that the degree of crystallization achieved with 400 nm-FLA is even higher than that of 800 nm-FLA. Cross-sectional scanning electron microscopic images indicate that the 100 nm-thick a-Si film is not fully crystallized. This can be explained by the much shorter penetration depth of 400 nm light than that of 800 nm light in a-Si.

Recently, deep neural network (DNN) schemes based on big data-driven methods have been successfully applied to image classification, communication, translation of language, speech recognition, etc. However, more efforts are still needed to apply them to complex robust nonlinear filter design in signal processing, especially for the robust nonlinear H ∞ filter design for robust state estimation of nonlinear stochastic signal system under uncertain external disturbance and output measurement noise. In general, the design problem of robust nonlinear H ∞ filter needs to solve a complex Hamilton-Jacobi-Isaacs equation (HJIE), which is not easily solved analytically or numerically. Further, robust nonlinear H ∞ filter is not easily designed by training DNN directly via conventional big data schemes. In this paper, a novel robust H ∞ HJIE-embedded DNN-based filter design is proposed as a co-design of H ∞ filtering algorithm and DNN learning algorithm for the robust state estimation of nonlinear stochastic signal systems with external disturbance and output measurement noise. In the proposed robust H ∞ DNN-based filter design, we have proven that when the approximation error of HJIE by the trained DNN through Adam learning algorithm approaches to 0, the HJIE-embedded DNN-based filter will approach the robust nonlinear H ∞ filter of nonlinear stochastic signal system with uncertain external disturbance and output measurement noise. Finally, a trajectory estimation problem of 3-D geometry incoming nonlinear stochastic missile system by the proposed robust H ∞ HJIE-embedded DNN-based filter scheme through the measurement by the sensor of radar system with external disturbance and measurement noise is given to illustrate the design procedure and validate its robust H ∞ filtering performance when compared with the extended Kalman filter and particle filter.INDEX TERMS Deep neural network (DNN), robust H ∞ filter, nonlinear stochastic signal system, extended Kalman filter, particle filter, Hamilton-Jacobi Isaacs equation (HJIE), co-design of H ∞ filtering and deep neural network learning.

No abstract

No abstract

No abstract