The finishing operation of the electrochemical finishing technology (ECF) for tube of steel was investigated In this study. Experimental procedures included qualitative
and quantitative analyses for surface roughness and material removal. Qualitative analyses utilized finishing optimization of a specific specimen in various design and operating conditions; value of gap from 0.2 to 10mm, flow rate of electrolytes from 5 to 15liter/min, finishing time from 1 to 4min and the applied voltage from 6 to 12v, to find out the value of surface roughness and material removal at each electrochemical state. From the measured material removal for each process state was used to verify the relationship with finishing time of work piece. Electrochemical finishing proves an effective method to reduce the surface roughness (Ra) from 1.6μm to 0.1μm in 4 min. Finally, the observed relationships were used to predicate the diameter of blank, tool diameter and flow rate by neural network modeling ANN which has inputs defined by the finished hole diameter, surface roughness, and finishing time. Three of hidden layers and their neurons were found by an integration procedure. The design charts observed from this study utilize the designers in predication of diameter for blank and design of electrode.
This paper introduces a Laplace-based modeling approach for the study of transient converter-grid interactions. The proposed approach is based on the development of two-port admittance models of converters and other components, combined with the use of numerical Laplace transforms. The application of a frequency domain method is aimed at the accurate and straightforward computation of transient system responses while preserving the wideband frequency characteristics of power components, such as those due to the use of high frequency semiconductive switches, electromagnetic interaction between inductive and capacitive components, as well as wave propagation and frequency dependence in transmission systems.
Imitation learning is an effective method for training an autonomous agent to accomplish a task by imitating expert behaviors in their demonstrations. However, traditional imitation learning methods require a large number of expert demonstrations in order to learn a complex behavior. Such a disadvantage has limited the potential of imitation learning in complex tasks where the expert demonstrations are not sufficient. In order to address the problem, we propose a Generative Adversarial Network-based model which is designed to learn optimal policies using only a single demonstration. The proposed model is evaluated on two simulated tasks in comparison with other methods. The results show that our proposed model is capable of completing co
... Show MoreThe blade pitch angle (BPA) in wind turbine (WT) is controlled to maximize output power generation above the rated wind speed (WS). In this paper, four types of controllers are suggested and compared for BPA controller in WT: PID controller (PIDC), type-1 fuzzy logic controller (T1-FLC), type-2 fuzzy logic controller (T2-FLC), and hybrid fuzzy-PID controller (FPIDC). The Mamdani and Sugeno fuzzy inference systems (FIS) have been compared to find the best inference system used in FLC. Genetic algorithm (GA) and Particle swarm optimization algorithm (PSO) are used to find the optimal tuning of the PID parameter. The results of500-kw horizontal-axis wind turbine show that PIDC based on PSO can reduced 2.81% in summation error of power
... Show MoreA simple, fast, inexpensive and sensitive method has been proposed to screen and optimize experimental factors that effecting the determination of phenylephrine hydrochloride (PHE.HCl) in pure and pharmaceutical formulations. The method is based on the development of brown-colored charge transfer (CT) complex with p-Bromanil (p-Br) in an alkaline medium (pH=9) with 1.07 min after heating at 80 °C. ‘Design of Experiments’ (DOE) employing ‘Central Composite Face Centered Design’ (CCF) and ‘Response Surface Methodology’ (RSM) were applied as an improvement to traditional ‘One Variable at Time’ (OVAT) approach to evaluate the effects of variations in selected factors (volume of 5×10-3 M p-Br, heating time, and temperature) on
... Show MoreThis research aims to solve the problem of selection using clustering algorithm, in this research optimal portfolio is formation using the single index model, and the real data are consisting from the stocks Iraqi Stock Exchange in the period 1/1/2007 to 31/12/2019. because the data series have missing values ,we used the two-stage missing value compensation method, the knowledge gap was inability the portfolio models to reduce The estimation error , inaccuracy of the cut-off rate and the Treynor ratio combine stocks into the portfolio that caused to decline in their performance, all these problems required employing clustering technic to data mining and regrouping it within clusters with similar characteristics to outperform the portfolio
... Show MoreThis study investigates the ionic conduction dependence on the size of alkaline cations in gel polymer electrolytes based on double iodide can enhance by incorporating a salt having a bulky cation.
... Show More