Training neural networks using Central Force Optimization and Particle Swarm Optimization: Insights and comparisons

Central Force Optimization (CFO) is a novel and upcoming metaheuristic technique that is based upon physical kinematics. It has previously been demonstrated that CFO is effective when compared with other metaheuristic techniques when applied to multiple benchmark problems and some real world applications. This work applies the CFO algorithm to training neural networks for data classification. As a proof of concept, the CFO algorithm is first applied to train a basic neural network that represents the logical XOR function. This work is then extended to train two different neural networks in order to properly classify members of the Iris data set. These results are compared and contrasted to results gathered using Particle Swarm Optimization (PSO) in the same applications. Similarities and differences between CFO and PSO are also explored in the areas of algorithm design, computational complexity, and natural basis. The paper concludes that CFO is a novel and promising meta-heuristic that is competitive with if not superior to the PSO algorithm, and there is much room to further improve it.     

Current and Temperature Dependent Matching Fields in Nb Film of Rectangular Array of Holes

Nanoengineered periodic array of holes on superconducting thin films have a great interest due to their excellence for the studies of the vortex pinning mechanisms in the type-II superconductors. Rectangular array of holes has been fabricated over a microbridge of Nb superconducting thin film by e-beam lithography. Rectangular array of holes have two type of scenario, at low magnetic fields matching effects are sharp and narrow while at high fields wide and shallower. In this work, we study the matching pinning effect by the artificial hole array in superconducting Nb thin films. We observed that as the inter distance between holes is decreased; the sharp matching effects become dominant and vice versa.     

Sampling-based online motion planning for mobile robots: utilization of Tabu search and adaptive neuro-fuzzy inference system

Despite the proven advantages of sampling-based motion planning algorithms, their inability to handle online navigation tasks and providing low-cost solutions make them less efficient in practice. In this paper, a novel sampling-based algorithm is proposed which is able to plan in an unknown environment and provides solutions with lower cost in terms of path length, runtime and stability of the results. First, a fuzzy controller is designed which incorporates the heuristic rules of Tabu search to enable the planner for solving online navigation tasks. Then, an adaptive neuro-fuzzy inference system (ANFIS) is proposed such that it constructs and optimizes the fuzzy controller based on a set of given input/output data. Furthermore, a heuristic dataset generator is implemented to provide enough data for the ANFIS using a randomized procedure. The performance of the proposed algorithm is evaluated through simulation in different motion planning queries. Finally, the proposed planner is compared to some of the similar motion planning algorithms to support the claim of superiority of its performance.

     

Development and Validation of an Artificial Mechanical Skin Model for the Study of Interactions between Skin and Microneedles

Microneedles are small needle‐like structures that are almost invisible to the naked eye. They have an immense potential to serve as a valuable tool in many medical applications, such as painless vaccination. Microneedles work by breaking through the stratum corneum, the outermost barrier layer of the skin, and providing a direct path for drug delivery into the skin. A lot of research has been presented over the past two decades on the applications of microneedles, yet the fundamental mechanism of how they interact, pressure, and penetrate the skin in its native state is worth examining further. As such, a major difficulty with understanding the mechanism of microneedle–skin interaction is the lack of an artificial mechanical human skin model to use as a standardized substrate. In this research news, the development of an artificial mechanical skin model based on a thorough mechanical study of fresh human and porcine skin samples is presented. The artificial mechanical skin model can be used to study the mechanical interactions between microneedles and skin, but not diffusion of molecules across skin. This model can assist in improving the performance of microneedles by enhancing the reproducibility of microneedle depth insertions for optimal drug delivery and biosensing.

     

Polymeric Nanoparticle-Based Photodynamic Therapy for Chronic Periodontitis in Vivo

Antimicrobial photodynamic therapy (aPDT) is increasingly being explored for treatment of periodontitis. Here, we investigated the effect of aPDT on human dental plaque bacteria in suspensions and biofilms in vitro using methylene blue (MB)-loaded poly(lactic-co-glycolic) (PLGA) nanoparticles (MB-NP) and red light at 660 nm. The effect of MB-NP-based aPDT was also evaluated in a clinical pilot study with 10 adult human subjects with chronic periodontitis. Dental plaque samples from human subjects were exposed to aPDT—in planktonic and biofilm phases—with MB or MB-NP (25 µg/mL) at 20 J/cm² in vitro. Patients were treated either with ultrasonic scaling and scaling and root planing (US + SRP) or ultrasonic scaling + SRP + aPDT with MB-NP (25 µg/mL and 20 J/cm²) in a split-mouth design. In biofilms, MB-NP eliminated approximately 25% more bacteria than free MB. The clinical study demonstrated the safety of aPDT. Both groups showed similar improvements of clinical parameters one month following treatments. However, at three months ultrasonic SRP + aPDT showed a greater effect (28.82%) on gingival bleeding index (GBI) compared to ultrasonic SRP. The utilization of PLGA nanoparticles encapsulated with MB may be a promising adjunct in antimicrobial periodontal treatment.     

Thermal Upgrading of Nickeliferous Pyrrhotite Tailings for the Recovery of Nickel in the Form of Ferronickel Alloy

Metallurgical and Materials Transactions B - Production of ferronickel alloy by thermal treatment of nickeliferous pyrrhotite (Pyrr) tailings was studied by both thermodynamic assessment and...     

Hydrothermal synthesis of different colors and morphologies of ZnO nanostructures and comparison of their photocatalytic properties

Various zinc oxide nanostructures were synthesized using thermal decomposition of zinc acetate dihydrate in a single process. The characterization of samples using powder X-ray diffraction, scanning electron microscope and FT-IR measurements revealed that the pure phase of different morphologies such as nanoparticles, nanowires and nanodisks had been synthesized successfully. Surprisingly some synthesized ZnO nanostructures were dark gray. The results showed that the reason may have been related to the oxygen deficiency and strong asymmetric stretching mode of wurtzite ZnO nanostructure. Using such samples, the photodegradation of Methylene blue was performed by UV–vis absorption measurement and the effect of morphology on the photocatalytic properties of different ZnO nanostructures was examined. The results showed that the nanodisks had the best photocatalytic performance among the other morphologies. The reason was attributed to the presence of specific crystal planes such as (0001) facets in nanodisks which can improve their photocatalytic performance.     

Thermo-mechanical and metallurgical aspects in friction stir processing of AZ31 Mg alloy—A numerical and experimental investigation

Friction stir processing of AZ31 Mg alloy was investigated by numerical modeling and experiments. A CFD based, fully coupled, 3D, thermo-mechanical model was built to better understand the effect of process parameters on temperature, material flow and strain rate. In order to account for material softening phenomena at elevated temperatures and extremely high strain rates that occur during the FSP process, experimentally measured peak temperatures were utilized to introduce a correction function in the flow stress constitutive relation. The numerical results showed that rotational speed as compared to translational speed had a more dominant effect on temperature field and strain rate. In addition, the asymmetric material flow around the tool axis caused higher peak temperature and strain rate on the advancing side (AS), while the material in the path of tool pin was swept around the retreating side (RS). FSP experiments confirmed peak temperatures measured at sheet surface near shoulder perimeter on AS were always higher than corresponding RS peak temperatures, under the selected range of process parameters. In addition to thermo-mechanical aspects, the metallurgical characteristics of FSP i.e. mainly the grain size evolution was studied by optical and electron microscopy. Experiments revealed that the coarse bimodal microstructure of as-received AZ31 Mg was subdivided into a defect-free, fine grain microstructure at the rotational speed of 1000 rpm, while a defect-free but a relatively coarse and bimodal microstructure evolved in the material at rotational speeds higher than 1000 rpm. Furthermore, in the selected range of process parameters the increases in translational speed resulted in finer grain sizes without the formation of voids or defects.     

Purification of metallurgical silicon using iron as an impurity getter part I: Growth and separation of Si

The purification of metallurgical grade silicon (MG-Si) using a combination of solvent refining and physical separation is studied. MG-Si was alloyed with iron and solidified under different cooling rates in order to grow pure Si dendrites from the alloy. The Si dendrites were then separated using a gravity-based method. The separation method relies on the significantly different densities of Si and FeSi₂, and it uses a heavy liquid with specific gravity between the two phases to float the light Si particles to the surface of the liquid, while the heavy iron silicide sinks. The effects of the particle size and cooling rate on the yield and separation efficiency of the Si phase were investigated by quantifying the fraction of Si in the sinks and floats. The results demonstrate that the crushing size of the particles prior to separation should be approximately the same as the width of the dendrites in order to maximize the separation efficiency while simultaneously lowering the grinding cost.