Kwong‐Sak Leung, Jian‐Yong Sun and Zong‐Ben Xu
In this paper, a set of safe adaptive genetic algorithms (sGAs) is proposed based on the Splicing/Decomposable encoding scheme and the efficient speed‐up strategies developed by Xu…
Abstract
In this paper, a set of safe adaptive genetic algorithms (sGAs) is proposed based on the Splicing/Decomposable encoding scheme and the efficient speed‐up strategies developed by Xu et al.. The proposed algorithms implement the self‐adaptation of the problem representation, selection and recombination operators at the levels of population, individual and component which commendably balance the conflicts between “reliability” and “efficiency”, as well as “exploitation” and “exploration” existed in the evolutionary algorithms. It is shown that the algorithms converge to the optimum solution in probability one. The proposed sGAs are experimentally compared with the classical genetic algorithm (CGA), non‐uniform genetic algorithm (nGA) proposed by Michalewicz, forking genetic algorithm (FGA) proposed by Tsutsui et al. and the classical evolution programming (CEP). The experiments indicate that the new algorithms perform much more efficiently than CGA and FGA do, comparable with the real‐coded GAs — nGA and CEP. All the algorithms are further evaluated through an application to a difficult real‐life application problem: the inverse problem of fractal encoding related to fractal image compression technique. The results for the sGA is better than those of CGA and FGA, and has the same, sometimes better performance compared to those of nGA and CEP.
Details
Keywords
Hui‐Yuan Fan, Jane Wei‐Zhen Lu and Zong‐Ben Xu
Genetic algorithms have been extensively used in different domains as a type of robust optimization method. They have a much better chance of achieving global optima than…
Abstract
Genetic algorithms have been extensively used in different domains as a type of robust optimization method. They have a much better chance of achieving global optima than conventional gradient‐based methods which usually converge to local sub‐optima. However, convergence speeds of genetic algorithms are often not good enough at their current stage. For this reason, improving the existing algorithms becomes a very important aspect of accelerating the development of the algorithms. Three improved strategies for genetic algorithms are proposed based on Holland’s simple genetic algorithm (SGA). The three resultant improved models are studied empirically and compared, in feasibility and performance evaluation, with a set of artificial test functions which are usually used as performance benchmarks for genetic algorithms. The simulation results demonstrate that the three proposed strategies can significantly improve the SGA.
Details
Keywords
Fashu Xu, Rui Huang, Hong Cheng, Min Fan and Jing Qiu
This paper aims at the problem of attaching the data of doctors, patients and the real-time sensor data of the exoskeleton to the cloud in intelligent rehabilitation applications…
Abstract
Purpose
This paper aims at the problem of attaching the data of doctors, patients and the real-time sensor data of the exoskeleton to the cloud in intelligent rehabilitation applications. This study designed the exoskeleton cloud-brain platform and validated its safety assessment.
Design/methodology/approach
According to the dimension of data and the transmission speed, this paper implements a three-layer cloud-brain platform of exoskeleton based on Alibaba Cloud's Lambda-like architecture. At the same time, given the human–machine safety status detection problem of the exoskeleton, this paper built a personalized machine-learning safety detection module for users with the multi-dimensional sensor data cloned by the cloud-brain platform. This module includes an abnormality detection model, prediction model and state classification model of the human–machine state.
Findings
These functions of the exoskeleton cloud-brain and the algorithms based on it were validated by the experiments, they meet the needs of use.
Originality/value
This thesis innovatively proposes a cloud-brain platform for exoskeletons, beginning the digitalization and intelligence of the exoskeletal rehabilitation process and laying the foundation for future intelligent assistance systems.