Search

Due to the unique characteristics of near-fault ground motions with an intense pulse, near-fault seismic sequences have the potential to cause more severe damage to structures compared to far-fault seismic sequences. This study focuses on quantitatively comparing the seismic response and fragility of a containment structure subjected to near-fault and far-fault seismic sequences. For this purpose, both as-recorded and artificial seismic sequences are used as input to conduct the nonlinear dynamic analysis. The effect of fault types of aftershocks on a mainshock-damaged containment is investigated in terms of the global response and local damage, respectively. Seismic demands on secondary systems are also studied through floor response spectra considering post-mainshock damage states. To evaluate the influence of near-fault and far-fault seismic sequences on the fragility of a containment structure, a new methodology is used to generate fragility curves, which can effectively consider the effect of post-mainshock damage states when computing the exceedance probability. The results show that near-fault seismic sequences can cause greater dynamic responses of a containment than far-fault ground motions, and the fragility evaluation without considering near-fault seismic records can overestimate the safety margin of a containment structure.

A loss-of-thrust UAV (unmanned aerial vehicle) needs to land to a predetermined area, which not only requires a higher accuracy of landing area, but also needs to maintain a certain attitude angle. In this paper, the Schwartz inequality and state space form are used to derive the guidance law with angle constraint. The influence of heading error and angle constraint on aircraft landing area are analyzed quantitatively based on the adjoint method. The simulations show the range of the heading error and angle constraint on the condition of certain miss distance. Compared with traditional Monte-Carlo method, the method of adjoint in this paper has the advantages of small computation and simple data processing.

Multilayer feedforward networks are related to functional groups of neurons where information is transmitted from one group to the next. It is a generic framework to characterize the properties of spiking activities in propagation. In this paper, we designed a novelty structure of feedforward network on FPGA to propagate the synchronous spiking activities. The experiment results demonstrate that the hardware implemented feedforward network can achieve the same synchronized states as the results of software simulation and show the high performance of FPGA implementation. The implementation of the feedforward network using one of the Altera Stratix III series FPGA boards is presented.

2 pp. (1997).