An improved Differential Evolution algorithm using learning automata and population topologies

A new variant of Differential Evolution (DE), called ADE-Grid, is presented in this paper which adapts the mutation strategy, crossover rate (CR) and scale factor (F) during the run. In ADE-Grid, learning automata (LA), which are powerful decision making machines, are used to determine the proper value of the parameters CR and F, and the suitable strategy for the construction of a mutant vector for each individual, adaptively. The proposed automata based DE is able to maintain the diversity among the individuals and encourage them to move toward several promising areas of the search space as well as the best found position. Numerical experiments are conducted on a set of twenty four well-known benchmark functions and one real-world engineering problem. The performance comparison between ADE-Grid and other state-of-the-art DE variants indicates that ADE-Grid is a viable approach for optimization. The results also show that the proposed ADE-Grid improves the performance of DE in terms of both convergence speed and quality of final solution.     

A note on the learning automata based algorithms for adaptive parameter selection in PSO

PSO, like many stochastic search methods, is very sensitive to efficient parameter setting such that modifying a single parameter may cause a considerable change in the result. In this paper, we study the ability of learning automata for adaptive PSO parameter selection. We introduced two classes of learning automata based algorithms for adaptive selection of value for inertia weight and acceleration coefficients. In the first class, particles of a swarm use the same parameter values adjusted by learning automata. In the second class, each particle has its own characteristics and sets its parameter values individually. In addition, for both classed of proposed algorithms, two approaches for changing value of the parameters has been applied. In first approach, named adventurous, value of a parameter is selected from a finite set while in the second approach, named conservative, value of a parameter either changes by a fixed amount or remains unchanged. Experimental results show that proposed learning automata based algorithms compared to other schemes such as SPSO, PSOIW, PSO-TVAC, PSOLP, DAPSO, GPSO, and DCPSO have the same or even higher ability to find better solutions. In addition, proposed algorithms converge to stopping criteria for some of the highly multi modal functions significantly faster.     

Speeding up learning automata based multi agent systems using the concepts of stigmergy and entropy

Learning automata (LA) were recently shown to be valuable tools for designing Multi-Agent Reinforcement Learning algorithms and are able to control the stochastic games. In this paper, the concepts of stigmergy and entropy are imported into learning automata based multi-agent systems with the purpose of providing a simple framework for interaction and coordination in multi-agent systems and speeding up the learning process. The multi-agent system considered in this paper is designed to find optimal policies in Markov games. We consider several dummy agents that walk around in the states of the environment, make local learning automaton active, and bring information so that the involved learning automaton can update their local state. The entropy of the probability vector for the learning automata of the next state is used to determine reward or penalty for the actions of learning automata. The experimental results have shown that in terms of the speed of reaching the optimal policy, the proposed algorithm has better learning performance than other learning algorithms.     

BNC-VLA: bayesian network structure learning using a team of variable-action set learning automata

Bayesian Network (BN) is a probabilistic graphical model which describes the joint probability distribution over a set of random variables. One of the most important challenges in the field of BNs is to find an optimal network structure based on an available training dataset. Since the problem of searching the optimal BN structure belongs to the class of NP-hard problems, typically greedy algorithms are used to solve it. In this paper a learning automata-based algorithm has been proposed to solve the BNs structure learning problem. There is a learning automaton corresponding with each random variable and at each stage of the proposed algorithm, named BNC-VLA, a set of learning automata is randomly activated and determined the graph edges that must be appeared in that stage. Finally, the constructed network is evaluated using a scoring function. As BNC-VLA algorithm proceeds, the learning process focuses on the BN structure with higher scores. The convergence of this algorithm is theoretically proved; and also some experiments are designed to evaluate the performance of it. Experimental results show that BNC-VLA is capable of finding the optimal structure of BN in an acceptable execution time; and comparing against other search-based methods, it outperforms them.     

Performance analysis of multi-antenna relay networks with imperfect channel estimation

In this paper, we present the performance of multi-antenna selective combining decode-and-forward (SC-DF) relay networks over independent but non-identical Nakagami-m fading channels with imperfect channel estimation. The outage probability, moment generating function (MGF) and symbol error probability (SEP) will be derived in closed-form using the SNR statistical characteristics. To make the analysis trackable, we have derived the MGF and SEP for integer values of fading severity, m. Also, to make the relations more simple, we develop high signal to noise ratio (SNR) analysis for the performance metrics of our system. Subsequently, we propose optimal and adaptive power allocation algorithms along with the equal power allocation method. Finally, for comparison with analytical formulas, we perform some Monte-Carlo simulations.     

Nanomechanical and tribological behavior of hydroxyapatite reinforced ultrahigh molecular weight polyethylene nanocomposites for biomedical applications

A hydroxyapatite (HA) particulate reinforced ultrahigh molecular weight polyethylene (UHMWPE) nanocomposite is fabricated by internal mixer at 180°C and using of paraffin oil as a processing aid to overcome the high viscosity of melted UHMWPE. The reinforcing effects of nano‐HA are investigated on nanomechanical properties of HA/UHMWPE nanocomposites by nanoindentation and nanoscratching methods. Results show that the nanocomposite with 50 wt % nano‐HA exhibits a Young's modulus and hardness of 362.5% and 200% higher, and a friction coefficient of 38.86% lower than that of pure UHMWPE, respectively. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42052.     

Fabrication of oriented electrospun cellulose nanocrystals–polystyrene composite fibers on a rotating drum

Nitrobenzene (CNC-1), trifluoromethyl benzene (CNC-2) modified and polystyrene-grafted (CNC-g) cellulose nanocrystals in polystyrene (PS)-N,N dimethylformamide (DMF) solutions were electrospun and collected as stretched and aligned fibers on a rotating drum. Scanning electron microscope pictures showed significant alignment in the case of unmodified and nitrobenzene-modified CNC-1/PS nanocomposite fibers once the linear speed of rotor reached to 15 m s⁻¹. Fiber diameter decrease was more strong with rotor speed increase in the case of trifluoromethyl benzene modified (CNC-2) and polystyrene-grafted (CNC-g) cellulose nanocrystals/PS systems. Dynamic mechanical analysis including storage and elastic modulus of electrospun-oriented fibers were performed on surface-modified and polymer-grafted CNC/PS samples. According to α transition peak, the increase in the glass-transition temperature with filler concentration was the highest in polymer-grafted CNC-g/PS composite fibers. It was due to the interpenetration of grafted polymer brushes and free polymer chains in continuous phase and resulted in restrictions of motions of polymer chains in the PS matrix. The elastic moduli of nitrobenzene (CNC-1) and trifluoromethyl benzene (CNC-2)-modified CNC-filled PS composite fibers agreed well with percolation model, which indicates the CNC–CNC interactions and network formation with an increase in concentration. Magnitude of the elastic modulus of polymer grafted CNC-g at 0.33 vol % in PS was significantly higher than the prediction from percolation theory. It was due the immobilized polymer chains around CNC-g particles. However, grafted polymer chains, at higher CNC concentrations acted like stickers among CNC particles and caused CNC agglomerates with entrapped free polystyrene from the matrix, thus caused a decrease in the elastic modulus. © 2020 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2020 , 137, 48942.     

LAMR: learning automata based multicast routing protocol for multi-channel multi-radio wireless mesh networks

Multicast routing is a crucial issue in wireless networks in which the same content should be delivered to a group of recipients simultaneously. Multicast is also considered as a key service for audio and video applications as well as data dissemination protocols over the last-mile backhaul Internet connectivity provided by multi-channel multi-radio wireless mesh networks (MCMR WMNs). The multicast problem is essentially related to a channel assignment strategy which determines the most suitable channel-radio associations. However, channel assignment brings about its own complications and hence, solving the multicast problem in MCMR WMNs will be more complicated than that of traditional networks. This problem has been proved to be NP-hard. In the major prior art multicast protocols developed for these networks, channel assignment and multicast routing are considered as two separate sub-problems to be solved sequentially. The work in this article is targeted at promoting the adoption of learning automata for joint channel assignment and multicast routing problem in MCMR WMNs. In the proposed scheme named LAMR, contrary to the existing methods, these two sub-problems will be solved conjointly. Experimental results demonstrate that LAMR outperforms the LCA and MCM proposed by Zeng et al. (IEEE Trans. Parallel. Distrib. Syst. 21(1):86–99, 2010) as well as the genetic algorithm-, tabu search-, and simulated annealing-based methods by Cheng and Yang (Int. J. Appl. Soft Comput. 11(2):1953–1964, 2011) in terms of achieved throughput, end-to-end delay, average packet delivery ratio, and multicast tree total cost.     

Impact of carbon tax on internal combustion engine size selection in a medium scale CHP system

Combined heat and power (CHP) systems due to their high efficiency compared to the conventional power generation systems have received considerable attention as they have less harmful impact on the environment. Recently, the serious concern with reducing the greenhouse gas emissions has focussed the attention on the possibility of a carbon tax in some countries. Here, we address the impact of such tax on the sizing and economics of a CHP system.     

System reliability analysis of soldier-piled excavation in unsaturated soil by combining random finite element and sequential compounding methods

Bulletin of Engineering Geology and the Environment - The stability of excavation in unsaturated soil is closely related to the variation of soil properties and matric suction. Determination of...