نویسندگان | Saeed Khorashadizadeh,Esmaeil Salahshour,Milad Malekzadeh,Reza Gholipour |
---|---|
نشریه | Evolving Systems |
شماره صفحات | 317-331 |
شماره سریال | 10 |
شماره مجلد | 3 |
نوع مقاله | Full Paper |
تاریخ انتشار | 2019 |
رتبه نشریه | علمی - مروری |
نوع نشریه | چاپی |
کشور محل چاپ | ایران |
نمایه نشریه | Scopus |
چکیده مقاله
Due to outstanding learning capabilities, neural networks have attracted a great deal of research interests and various structures for them have been presented. Quantum neural network is one these structures used in quantum calculations to improve learning capabilities of neural networks. In this paper, using a quantum neural network, a chaos controller for rod-type plasma torch system is developed. This system is extensively nonlinear and its dynamics shows considerable sensitivity to its parameters that makes it a good challenge for investigation of controllers’ performances. It is very common to utilize particle swarm optimization (PSO) algorithm to determine the parameters of neural networks. The novelty of this paper is changing the search space of the parameters in each particle based on that parameter’s value. Also, weaker particles will be deleted and new particles will be substituted. These modifications have enhanced the proposed algorithm performance by adding the ability of global search and also searching in a reasonable space. To highlight the superiority of the proposed algorithm, the results of eight meta-heuristics optimization algorithms such as artificial bee colony, social spider algorithm and imperialist competitive algorithm are presented for comparison. Simulation results show that the improved PSO (IPSO) proposed in this paper is superior to other algorithms from the viewpoint of convergence speed and accuracy. Moreover, in chaos control of a rod-type plasma torch system, the proposed method outperforms a PID controller and also a neural network controller using perceptron.
tags: Quantum neural network · Improved particle swarm optimization · Rod-type plasma torch system · Chaos control