A Novel Machine-Learning Aided Optimization Technique for Material Design: Application in Thin Film Solar Cells

2016 ◽  
Author(s):  
Shima Hajimirza
Keyword(s):  
Thin Film ◽  
Machine Learning ◽  
Objective Function ◽  
Search Algorithm ◽  
Random Search ◽  
Optimization Technique ◽  
Global Search ◽  
Radiative Properties ◽  
Mcmc Methods ◽  
Predictive Tool

Patterned thin film structures can offer spectrally selective radiative properties that benefit many engineering applications including photovoltaic energy conversion at extremely efficient scales. Inverse design of such structures can be expressed as an interesting optimization problem with a specific regime of complexity; namely moderate number of optimization parameters but highly time-consuming forward problem. For problems like this, a search technique that can somehow learn and parameterize the multi-dimensional behavior of the objective function based on past search points can be extremely useful in guiding the global search algorithm and expediting the solution for such complexity regimes. Based on this idea, we have developed a novel search algorithm for optimizing absorption coefficient of visible light in a multi-layered silicon-based nano-scale thin film solar cell. The proposed optimization algorithm uses a machine-learning predictive tool called regression-tree in an intermediary step to learn (i.e. regress) the objective function based on a previous generation of random search points. The fitted model is then used as a guide to resample from a new generation of candidate solutions with a significantly higher average gain. This process can be repeated multiple times and better solutions are obtained with high likelihood at each stage. Through numerical experiments we demonstrate how in only one resampling stage, the propose technique dominates the state-of-the-art global search algorithms such as gradient based techniques or MCMC methods in the considered nano-design problem.

scholarly journals Genetic Algorithm – A Sensible Evolutionary Optimization Technique

2019 ◽  
Vol 8 (9) ◽  
pp. 2805-2809
Keyword(s):  
Genetic Algorithm ◽  
Optimization Problems ◽  
Search Algorithm ◽  
Random Search ◽  
Optimization Technique ◽  
Search Space ◽  
Effective Population ◽  
Natural Systems ◽  
Heuristic Search Algorithm ◽  
Feasible Solutions

The study presents a pragmatic outlook of genetic algorithm. Many biological algorithms are inspired for their ability to evolve towards best solutions and of all; genetic algorithm is widely accepted as they well suit evolutionary computing models. Genetic algorithm could generate optimal solutions on random as well as deterministic problems. Genetic algorithm is a mathematical approach to imitate the processes studied in natural evolution. The methodology of genetic algorithm is intensively experimented in order to use the power of evolution to solve optimization problems. Genetic algorithm is an adaptive heuristic search algorithm based on the evolutionary ideas of genetics and natural selection. Genetic algorithm exploits random search approach to solve optimization problems. Genetic algorithm takes benefits of historical information to direct the search into the convergence of better performance within the search space. The basic techniques of evolutionary algorithms are observed to be simulating the processes in natural systems. These techniques are aimed to carry effective population to the next generation and ensure the survival of the fittest. Nature supports the domination of stronger over the weaker ones in any kind. In this study, we proposed the arithmetic views of the behavior and operators of genetic algorithm that support the evolution of feasible solutions to optimized solutions.

A Tabu-Gradient Search-Based Optimization Technique for Synthesis of Mechanisms

2004 ◽  
Author(s):  
Ahmad Smaili ◽  
Naji Atallah
Keyword(s):  
Tabu Search ◽  
Global Minimum ◽  
Short Term Memory ◽  
Search Algorithm ◽  
Random Search ◽  
Optimization Technique ◽  
Optimization Methods ◽  
Gradient Search ◽  
Optimum Synthesis ◽  
Timing Task

Mechanism synthesis requires the use of optimization methods to obtain approximate solution whenever the desired number of positions the mechanism is required to traverse exceeds a few (five in a 4R linkage). Deterministic gradient-based methods are usually impractical when used alone because they move in the direction of local minima. Random search methods on the other hand have a better chance of converging to a global minimum. This paper presents a tabu-gradient search based method for optimum synthesis of planar mechanisms. Using recency-based short-term memory strategy, tabu-search is initially used to find a solution near global minimum, followed by a gradient search to move the solution ever closer to the global minimum. A brief review of tabu search method is presented. Then, tabu-gradient search algorithm is applied to synthesize a four-bar mechanism for a 10-point path generation with prescribed timing task. As expected, Tabu-gradient base search resulted in a better solution with less number of iterations and shorter run-time.

Diagnostic Strategy Optimization Based on Predatory Search Algorithm

2012 ◽  
Vol 591-593 ◽  
pp. 2441-2444
Author(s):  
Jin Luo ◽  
Qi Bin Deng ◽  
Chen Meng
Keyword(s):  
Local Search ◽  
Objective Function ◽  
Search Algorithm ◽  
Optimal Solution ◽  
Search Space ◽  
Global Search ◽  
State Probability ◽  
Diagnostic Strategy ◽  
Local Optimal Solution ◽  
Predatory Search Algorithm

With respect to the inherent NP-hard complexity of Optimization of testability diagnostic strategy problem, a predatory search algorithm simulating animal predatory strategies was designed. This algorithm adopted the gross test expense including state probability, isolation matrix and test expense as its objective function, defined local and global search by the restriction value of search space based on two points exchange, and realized the conversion between local and global search by adjusting the restriction value of search space. It had better ability to conduct local search and jump out of local optimal solution simultaneously, and provided a better resolution for the optimization of testability diagnostic strategy.

Discrete-Continuous Configuration Optimization Methods for Structures Using the Harmony Search Algorithm

2006 ◽  
Vol 324-325 ◽  
pp. 1293-1296 ◽  
Author(s):  
K.S. Lee ◽  
Chang Sik Choi
Keyword(s):  
Search Algorithm ◽  
Random Search ◽  
Optimization Technique ◽  
Harmony Search ◽  
Optimization Methods ◽  
Harmony Search Algorithm ◽  
Perfect State ◽  
Gradient Search ◽  
Geometric Variables ◽  
Derivative Information

This paper proposes an efficient structural optimization methods based on the harmony search (HS) heuristic algorithm that treat integrated discrete sizing and continuous geometric variables. The recently developed HS algorithm was conceptualized using the musical process of searching for a perfect state of harmony. It uses a stochastic random search instead of a gradient search so derivative information is unnecessary. A benchmark truss example is presented to demonstrate the effectiveness and robustness of the new method, as compared to current optimization methods. The numerical results reveal that the proposed method is a powerful search and design optimization technique for structures with discrete member sizes, and may yield better solutions than those obtained using current methods.

scholarly journals A New Method of Secure Authentication Based on Electromagnetic Signatures of Chipless RFID Tags and Machine Learning Approaches

Sensors ◽  
2020 ◽  
Vol 20 (21) ◽  
pp. 6385
Author(s):  
Dragoș Nastasiu ◽  
Răzvan Scripcaru ◽  
Angela Digulescu ◽  
Cornel Ioana ◽  
Raymundo De Amorim ◽  
...  
Keyword(s):  
Machine Learning ◽  
Radio Frequency Identification ◽  
Search Algorithm ◽  
Random Search ◽  
Recognition Rate ◽  
Rfid Tags ◽  
Learning Approaches ◽  
V Band ◽  
Chipless Rfid ◽  
Secure Authentication

In this study, we present the implementation of a neural network model capable of classifying radio frequency identification (RFID) tags based on their electromagnetic (EM) signature for authentication applications. One important application of the chipless RFID addresses the counterfeiting threat for manufacturers. The goal is to design and implement chipless RFID tags that possess a unique and unclonable fingerprint to authenticate objects. As EM characteristics are employed, these fingerprints cannot be easily spoofed. A set of 18 tags operating in V band (65–72 GHz) was designed and measured. V band is more sensitive to dimensional variations compared to other applications at lower frequencies, thus it is suitable to highlight the differences between the EM signatures. Machine learning (ML) approaches are used to characterize and classify the 18 EM responses in order to validate the authentication method. The proposed supervised method reached a maximum recognition rate of 100%, surpassing in terms of accuracy most of RFID fingerprinting related work. To determine the best network configuration, we used a random search algorithm. Further tuning was conducted by comparing the results of different learning algorithms in terms of accuracy and loss.

scholarly journals Classification of Black Plastics Waste Using Fluorescence Imaging and Machine Learning

Recycling ◽  
2019 ◽  
Vol 4 (4) ◽  
pp. 40 ◽  
Author(s):  
Florian Gruber ◽  
Wulf Grählert ◽  
Philipp Wollmann ◽  
Stefan Kaskel
Keyword(s):  
Machine Learning ◽  
Classification Accuracy ◽  
Near Infrared ◽  
Search Algorithm ◽  
Random Search ◽  
Support Vector ◽  
Classification Algorithms ◽  
Linear Discriminant ◽  
Black Plastic

This work contributes to the recycling of technical black plastic particles, for example from the automotive or electronics industries. These plastics cannot yet be sorted with sufficient purity (up to 99.9%), which often makes economical recycling impossible. As a solution to this problem, imaging fluorescence spectroscopy with additional illumination in the near infrared spectral range in combination with classification by machine learning or deep learning classification algorithms is here investigated. The algorithms used are linear discriminant analysis (LDA), k-nearest neighbour classification (kNN), support vector machines (SVM), ensemble models with decision trees (ENSEMBLE), and convolutional neural networks (CNNs). The CNNs in particular attempt to increase overall classification accuracy by taking into account the shape of the plastic particles. In addition, the automatic optimization of the hyperparameters of the classification algorithms by the random search algorithm was investigated. The aim was to increase the accuracy of the classification models. About 400 particles each of 14 plastics from 12 plastic classes were examined. An attempt was made to train an overall model for the classification of all 12 plastics. The CNNs achieved the highest overall classification accuracy with 93.5%. Another attempt was made to classify 41 mixtures of industrially relevant plastics with a maximum of three plastic classes per mixture. The same average classification accuracy of 99.0% was achieved for the ENSEMBLE, SVM, and CNN algorithms. The target overall classification accuracy of 99.9% was achieved for 18 of the 41 compounds. The results show that the method presented is a promising approach for sorting black technical plastic waste.

IDENTIFY OF FRAUDULENT FINANCIAL OPERATIONS USING THE MACHINE LEARNING ALGORITHM

2020 ◽  
pp. 23-31
Author(s):  
S. L. Belyakov ◽  
S. М. Karpov
Keyword(s):  
Machine Learning ◽  
Learning Algorithm ◽  
Search Algorithm ◽  
Random Search ◽  
Experimental Comparison ◽  
Unbalanced Data ◽  
Ensemble Classifiers ◽  
Adaptive Boosting ◽  
Financial Transactions ◽  
Ensemble Algorithms

Current work is devoted to the problem of automatic detection of fraudulent financial transactions. The article describes the causes of fraudulent transactions their typical attributes, as well as the basic principle of detection. The concepts of fraudulent and honest transactions are defined. Examples of algorithms for determining suspicious financial transactions in antifraud systems are given. Modern approaches to monitoring and detecting cases of fraud in remote banking systems are considered. The positive and negative aspects of each approach are described. Particular attention is paid to the problem of optimal recognition of transaction classes in highly unbalanced data. Methods for solving the problem of unbalanced data are considered. The choice of means for evaluating the operation of the machine learning model is justified considering the specifics of data distribution. As a solution, we propose an approach based on the use of ensemble classifiers in conjunction with balanced sampling algorithms, the key feature of which is to create a balanced sample not for the entire classifier, but for each student in the ensemble separately. Based on data on fraud in the field of bank credit cards, a comparison is made and the best classifier is selected among such ensemble algorithms as random forest, adaptive boosting and bagging of decision trees. To create balanced subsets of evaluators of ensemble algorithms, the algorithm of random insufficient sampling is used. To search for the optimal parameters of the classifiers, the random search algorithm on the grid is used. The results of experimental comparison of the selected methods are presented. The advantages of the proposed approach are analyzed, and the boundaries of its applicability are discussed.

scholarly journals Economic and Ecological Design of Hybrid Renewable Energy Systems Based on a Developed IWO/BSA Algorithm

Electronics ◽  
2021 ◽  
Vol 10 (6) ◽  
pp. 687
Author(s):  
Mohammed Kharrich ◽  
Salah Kamel ◽  
Rachid Ellaia ◽  
Mohammed Akherraz ◽  
Ali S. Alghamdi ◽  
...  
Keyword(s):  
Optimal Design ◽  
Wind Turbine ◽  
Objective Function ◽  
Optimization Algorithm ◽  
Search Algorithm ◽  
Optimization Technique ◽  
Optimization Methods ◽  
Invasive Weed ◽  
Battery Capacity ◽  
Battery Systems

In this paper, an optimal design of a microgrid including four houses in Dakhla city (Morocco) is proposed. To make this study comprehensive and applicable to any hybrid system, each house has a different configuration of renewable energies. The configurations of these four houses are PV/wind turbine (WT)/biomass/battery, PV/biomass, PV/diesel/battery, and WT/diesel/battery systems. The comparison factor among these configurations is the cost of energy (COE), comparative index, where the load is different in the four houses. Otherwise, the main objective function is the minimization of the net present cost (NPC), subject to several operating constraints, the power loss, the power generated by the renewable sources (renewable fraction), and the availability. This objective function is achieved using a developed optimization algorithm. The main contribution of this paper is to propose and apply a new optimization technique for the optimal design of a microgrid considering different economic and ecological aspects. The developed optimization algorithm is based on the hybridization of two metaheuristic algorithms, the invasive weed optimization (IWO) and backtracking search algorithm (BSA), with the aim of collecting the advantages of both. The proposed hybrid optimization algorithm (IWO/BSA) is compared with the original two optimization methods (IWO and BSA) as well as other well-known optimization methods. The results indicate that PV/biomass and PV/diesel/battery systems have the best energy cost using the proposed IWO/BSA algorithm with 0.1184 $/kWh and 0.1354 $/kWh, respectively. The best system based on its LCOE factor is the PV/biomass which represents an NPC of 124,689 $, the size of this system is 349.55 m2 of PV area and the capacity of the biomass is 18.99 ton/year. The PV/diesel/battery option has also good results, with a system NPC of 142,233 $, the size of this system is about 391.39 m2 of PV area, rated power of diesel generator about 0.55 kW, and a battery capacity of 12.97 kWh. Otherwise, the proposed IWO/BSA has the best convergence in all cases. It is observed that the wind turbine generates more dumped power, and the PV system is highly suitable for the studied area.

Voltage Stability Enhancement with Optimal Placement of IPFC

2019 ◽  
Vol 8 (4) ◽  
pp. 9465-9471
Keyword(s):  
Transmission Line ◽  
Objective Function ◽  
Search Algorithm ◽  
Cuckoo Search ◽  
Test System ◽  
Optimal Location ◽  
Optimal Placement ◽  
Utilization Factor ◽  
Location Selection ◽  
Multi Objective

This paper presents a novel technique based on Cuckoo Search Algorithm (CSA) for enhancing the performance of multiline transmission network to reduce congestion in transmission line to huge level. Optimal location selection of IPFC is done using subtracting line utilization factor (SLUF) and CSA-based optimal tuning. The multi objective function consists of real power loss, security margin, bus voltage limit violation and capacity of installed IPFC. The multi objective function is tuned by CSA and the optimal location for minimizing transmission line congestion is obtained. The simulation is performed using MATLAB for IEEE 30-bus test system. The performance of CSA has been considered for various loading conditions. Results shows that the proposed CSA technique performs better by optimal location of IPFC while maintaining power system performance

scholarly journals Optimal Integration of Photovoltaic Sources in Distribution Networks for Daily Energy Losses Minimization Using the Vortex Search Algorithm

2021 ◽  
Vol 11 (10) ◽  
pp. 4418
Author(s):  
Alejandra Paz-Rodríguez ◽  
Juan Felipe Castro-Ordoñez ◽  
Oscar Danilo Montoya ◽  
Diego Armando Giral-Ramírez
Keyword(s):  
Objective Function ◽  
Power Flow ◽  
Search Algorithm ◽  
Peak Load ◽  
Distribution Networks ◽  
Energy Losses ◽  
Continuous Version ◽  
Continuous Section ◽  
Daily Energy ◽  
Discrete Part

This paper deals with the optimal siting and sizing problem of photovoltaic (PV) generators in electrical distribution networks considering daily load and generation profiles. It proposes the discrete-continuous version of the vortex search algorithm (DCVSA) to locate and size the PV sources where the discrete part of the codification defines the nodes. Renewable generators are installed in these nodes, and the continuous section determines their optimal sizes. In addition, through the successive approximation power flow method, the objective function of the optimization model is obtained. This objective function is related to the minimization of the daily energy losses. This method allows determining the power losses in each period for each renewable generation input provided by the DCVSA (i.e., location and sizing of the PV sources). Numerical validations in the IEEE 33- and IEEE 69-bus systems demonstrate that: (i) the proposed DCVSA finds the optimal global solution for both test feeders when the location and size of the PV generators are explored, considering the peak load scenario. (ii) In the case of the daily operative scenario, the total reduction of energy losses for both test feeders are 23.3643% and 24.3863%, respectively; and (iii) the DCVSA presents a better numerical performance regarding the objective function value when compared with the BONMIN solver in the GAMS software, which demonstrates the effectiveness and robustness of the proposed master-slave optimization algorithm.

Sign in / Sign up

Export Citation Format

Share Document