Strain measurement in a Mach–Zehnder fiber interferometer using genetic algorithm

Minimization of sensing elements for full-range optical pH device formulation

“…Here, neural networks 16 have been proposed to solve this problem, although the limitations of their classic training methods, becoming trapped in local optima solutions, 17 may make the model calibration processes difficult. Genetic algorithms 18 have proven to overcome these problems in some applications regarding sensor modeling, such as the generation of regional maps of a-chlorophyll concentration or total suspended matter from the ocean color, 19 strain sensing using a Mach-Zehnder fiber interferometer, 20,21 and strain and temperature sensing by fiber Bragg gratings. [22][23][24] One more recent training strategy for neural networks consists of the use of hybrid techniques that combine evolutionary algorithms with non-linear optimization methods to overcome local optima solutions.…”

Inverse modeling and control of a dielectric electro-active polymer smart actuator

“…In the literature, several studies successfully used a genetic algorithm (GA) to identify and optimize some typical nonlinear dynamic systems [24][25][26][27][28], especially for the parameters identification of the fuzzy systems, where the estimation of the rule base usually requires global searching due to the existence of if -then rules. However, in this paper, the APSO algorithm is applied to identify the remaining unknown parameters of the proposed inverse NARX fuzzy model.…”

Improving the Accuracy of Interference Microscope Measurements by Applying Genetic Algorithm

“…A large number of researches interested in using AI in surface metrology has been issued such as designing inference engine for correlating surface texture parameters with functional measures [3], Genetic algorithm (GA) was used in phase recovery from the interferogram with minimum phase ambiguities [4], also different clustering techniques such as k-means, ISODATA and neural networks were used to relate surface metrology data to a component's function and the manufacturing process that produced the sample [5].…”