Renewable energy (RE) resources are relatively unpredictable and dependent on climatic conditions. The negative effects of existing randomness in RE resources can be reduced by the integration of RE resources into what is called Hybrid Renewable Energy Systems (HRES). The design of HRES remains as a complicated problem since there is uncertainty in energy prices, demand, and RE sources. In addition, it is a multi-objective design since several conflicting objectives must be considered. In this thesis, an optimal sizing approach has been proposed to aid decision makers in sizing and performance analysis of this kind of energy supply systems. First, a straightforward methodology based on ε-constraint method is proposed for optimal sizing of HRESs containing RE power generators and two storage devices. The ε-constraint method has been applied to minimize simultaneously the total net present cost of the system, unmet load, and fuel emission. A simulation-based particle swarm optimization approach has been used to tackle the multi-objective optimization problem. In the next step, a Pareto-based search technique, named dynamic multi-objective particle swarm optimization, has been performed to improve the quality of the Pareto front (PF) approximated by the ε-constraint method. The proposed method is examined for a case study including wind turbines, photovoltaic panels, diesel generators, batteries, fuel cells, electrolyzers, and hydrogen tanks. Well-known metrics from the literature are used to evaluate the generated PF. Afterward, a multi-objective approach is presented to consider the economic, reliability and environmental issues at various renewable energy ratio values when optimizing the design of building energy supply systems. An existing commercial apartment building operating in a cold Canadian climate has been described to apply the proposed model. In this test application, the model investigates the potential use of RE resources for the building. Furthermore, the application of plug-in electric vehicles instead of gasoline car for transportation is studied. Comparing model results against two well-known reported multi-objective algorithms has also been examined. Finally, the existing uncertainties in RE and load are explicitly incorporated into the model to give more accurate and realistic results. An innovative and easy to implement stochastic multi-objective approach is introduced for optimal sizing of an HRES. / February 2016

AN ABSTRACT OF THE THESIS OF Mohamed A. Abuella, for the Master of Science degree in Electrical and Computer Engineering, presented on May 10, 2012, at Southern Illinois University Carbondale. TITLE:STUDY OF PARTICLE SWARM FOR OPTIMAL POWER FLOW IN IEEE BENCHMARK SYSTEMS INCLUDING WIND POWER GENERATORS MAJOR PROFESSOR: Dr. C. Hatziadoniu, The aim of this thesis is the optimal economic dispatch of real power in systems that include wind power. The economic dispatch of wind power units is quite different of conventional thermal units. In addition, the consideration should take the intermittency nature of wind speed and operating constraints as well. Therefore, this thesis uses a model that considers the aforementioned considerations in addition to whether the utility owns wind turbines or not. The optimal power flow (OPF) is solved by using one of the modern optimization algorithms: the particle swarm optimization algorithm (PSO). IEEE 30-bus test system has been adapted to study the implementation PSO algorithm in OPF of conventional-thermal generators. A small and simple 6-bus system has been used to study OPF of a system that includes wind-powered generators besides to thermal generators. The analysis of investigations on power systems is presented in tabulated and illustrative methods to lead to clear conclusions.

Population-based global optimization algorithms including Particle Swarm Optimization (PSO) have become popular for solving multi-optima problems much more efficiently than the traditional mathematical techniques. In this research, we present and evaluate a new parallel PSO algorithm that provides a significant performance improvement as compared to the serial PSO algorithm. Instead of merely assigning parts of the task of serial version to several processors, the new algorithm places multiple swarms on the available nodes in which operate independently, while collaborating on the same task. With the reduction of the communication bottleneck as well the ability to manipulate the individual swarms independently, the proposed approach outperforms the original PSO algorithm and still maintains the simplicity and ease of implementation.

Submitted by Maykon Nascimento (maykon.albani@hotmail.com) on 2015-08-03T18:48:30Z No. of bitstreams: 2 license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Um algoritmo PSO híbrido para planejamento de caminhos em navegação de veículos utilizando A.pdf: 2604695 bytes, checksum: ed8f69e49eaefe272bccd6025290c381 (MD5) / Approved for entry into archive by Elizabete Silva (elizabete.silva@ufes.br) on 2015-08-13T21:44:43Z (GMT) No. of bitstreams: 2 license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Um algoritmo PSO híbrido para planejamento de caminhos em navegação de veículos utilizando A.pdf: 2604695 bytes, checksum: ed8f69e49eaefe272bccd6025290c381 (MD5) / Made available in DSpace on 2015-08-13T21:44:43Z (GMT). No. of bitstreams: 2 license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Um algoritmo PSO híbrido para planejamento de caminhos em navegação de veículos utilizando A.pdf: 2604695 bytes, checksum: ed8f69e49eaefe272bccd6025290c381 (MD5) Previous issue date: 2015 / Utilizar robôs autônomos capazes de planejar o seu caminho é um desafio que atrai vários pesquisadores na área de navegação de robôs. Neste contexto, este trabalho tem como objetivo implementar um algoritmo PSO híbrido para o planejamento de caminhos em ambientes estáticos para veículos holonômicos e não holonômicos. O algoritmo proposto possui duas fases: a primeira utiliza o algoritmo A* para encontrar uma trajetória inicial viável que o algoritmo PSO otimiza na segunda fase. Por fim, uma fase de pós planejamento pode ser aplicada no caminho a fim de adaptá-lo às restrições cinemáticas do veículo não holonômico. O modelo Ackerman foi considerado para os experimentos. O ambiente de simulação de robótica CARMEN (Carnegie Mellon Robot Navigation Toolkit) foi utilizado para realização de todos os experimentos computacionais considerando cinco instâncias de mapas geradas artificialmente com obstáculos. O desempenho do algoritmo desenvolvido, A*PSO, foi comparado com os algoritmos A*, PSO convencional e A* Estado Híbrido. A análise dos resultados indicou que o algoritmo A*PSO híbrido desenvolvido superou em qualidade de solução o PSO convencional. Apesar de ter encontrado melhores soluções em 40% das instâncias quando comparado com o A*, o A*PSO apresentou trajetórias com menos pontos de guinada. Investigando os resultados obtidos para o modelo não holonômico, o A*PSO obteve caminhos maiores entretanto mais suaves e seguros. / Autonomous robots with the ability of planning their own way is a challenge that attracts many researchers in the area of robot navigation. In this context, this work aims to implement a hybrid PSO algorithm for planning paths in static environments for holonomic and non-holonomic vehicles. The proposed algorithm has two phases: the first uses A* algorithm to generates an initial and feasible trajectory which is optimized by the PSO algorithm in the second stage. Finally a post path planning phase can be applied in order to adapt it to non-holonomic vehicle kinematic constraints. The Ackerman model has been considered for the experiments. The Carnegie Mellon Robot Navigation Toolkit (CARMEN) was used to perform the computational experiments considering five instances of maps artificially generated with obstacles. The performance of the A*PSO algorithm was compared with A*, PSO and A*-Hybrid State. The results of the dynamic instances were not compared with other algorithms. The computational results indicates that the algorithm A*PSO outperformes the PSO algorithm. With respect to the algorithm A*, the A*PSO achieved better solutions for 40% of the tested instances, but all of them, with less waypoints. For non-holonomic instances, the A*PSO obtained longer paths, however smoother and safer.

abstract: In this era of fast computational machines and new optimization algorithms, there have been great advances in Experimental Designs. We focus our research on design issues in generalized linear models (GLMs) and functional magnetic resonance imaging(fMRI). The first part of our research is on tackling the challenging problem of constructing exact designs for GLMs, that are robust against parameter, link and model uncertainties by improving an existing algorithm and providing a new one, based on using a continuous particle swarm optimization (PSO) and spectral clustering. The proposed algorithm is sufficiently versatile to accomodate most popular design selection criteria, and we concentrate on providing robust designs for GLMs, using the D and A optimality criterion. The second part of our research is on providing an algorithm that is a faster alternative to a recently proposed genetic algorithm (GA) to construct optimal designs for fMRI studies. Our algorithm is built upon a discrete version of the PSO. / Dissertation/Thesis / Doctoral Dissertation Statistics 2014

Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2016-06-10T00:04:44Z No. of bitstreams: 1 AndreFelipeOliveiraDeAzevedoDantas_TESE.pdf: 1568997 bytes, checksum: 5540f3a6014383c6e5ad913d025e8220 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2016-06-10T22:40:15Z (GMT) No. of bitstreams: 1 AndreFelipeOliveiraDeAzevedoDantas_TESE.pdf: 1568997 bytes, checksum: 5540f3a6014383c6e5ad913d025e8220 (MD5) / Made available in DSpace on 2016-06-10T22:40:15Z (GMT). No. of bitstreams: 1 AndreFelipeOliveiraDeAzevedoDantas_TESE.pdf: 1568997 bytes, checksum: 5540f3a6014383c6e5ad913d025e8220 (MD5) Previous issue date: 2015-07-06 / Recentemente diversas t?cnicas de computa??o evolucion?rias t?m sido utilizadas em ?reas como estima??o de par?metros de processos din?micos lineares e n?o lineares ou at? sujeitos a incertezas. Isso motiva a utiliza??o de algoritmos como o otimizador por nuvem de part?culas (PSO) nas referidas ?reas do conhecimento. Por?m, pouco se sabe sobre a converg?ncia desse algoritmo e, principalmente, as an?lises e estudos realizados t?m se concentrado em resultados experimentais. Por isso, ? objetivo deste trabalho propor uma nova estrutura para o PSO que permita analisar melhor a converg?ncia do algoritmo de forma anal?tica. Para isso, o PSO ? reestruturado para assumir uma forma matricial e reformulado como um sistema linear por partes. As partes ser?o analisadas de forma separada e ser? proposta a inser??o de um fator de esquecimento que garante que a parte mais significativa deste sistema possua autovalores dentro do c?rculo de raio unit?rio. Tamb?m ser? realizada a an?lise da converg?ncia do algoritmo como um todo, utilizando um crit?rio de converg?ncia quase certa, aplic?vel a sistemas chaveados. Na sequ?ncia, ser?o realizados testes experimentais de maneira a verificar o comportamento dos autovalores ap?s a inser??o do fator de esquecimento. Posteriormente, os algoritmos de identifica??o de par?metros tradicionais ser?o combinados com o PSO matricial, de maneira a tornar os resultados da identifica??o t?o bons ou melhores que a identifica??o apenas com o PSO ou, apenas com os algoritmos tradicionais. Os resultados mostram a converg?ncia das part?culas em uma regi?o delimitada e que as fun??es obtidas ap?s a combina??o do algoritmo PSO matricial com os algoritmos convencionais, apresentam maior generaliza??o para o sistema apresentado. As conclus?es a que se chega ? que a hibridiza??o, apesar de limitar a busca por uma part?cula mais apta do PSO, permite um desempenho m?nimo para o algoritmo e ainda possibilita melhorar o resultado obtido com os algoritmos tradicionais, permitindo a representa??o do sistema aproximado em quantidades maiores de frequ?ncias.

Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2017-01-27T12:26:53Z No. of bitstreams: 1 FabioAugustoProcopioDePaiva_TESE.pdf: 11863653 bytes, checksum: ea2b87d3ec0832aff7e2d5c1c7eda033 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2017-01-30T12:32:55Z (GMT) No. of bitstreams: 1 FabioAugustoProcopioDePaiva_TESE.pdf: 11863653 bytes, checksum: ea2b87d3ec0832aff7e2d5c1c7eda033 (MD5) / Made available in DSpace on 2017-01-30T12:32:55Z (GMT). No. of bitstreams: 1 FabioAugustoProcopioDePaiva_TESE.pdf: 11863653 bytes, checksum: ea2b87d3ec0832aff7e2d5c1c7eda033 (MD5) Previous issue date: 2016-07-01 / Em muitos problemas de engenharia, ? comum o estudo de um tipo de processo que se comporta, via de regra, como um sistema din?mico. Esse tipo de sistema possui a peculiaridade de poder ser modelado por meio de um conjunto de equa??es que evolui ao longo do tempo para representar o comportamento modelado do sistema. Para resolver esses problemas de engenharia, diversos m?todos de Computa??o Bio-inspirada v?m sendo propostos como solu??o em diferentes contextos. Entre esses m?todos, est? uma categoria de algoritmos conhecida como Intelig?ncia de Enxames. Apesar do relativo sucesso, a maioria dos m?todos bio-inspirados enfrenta um problema muito comum conhecido como converg?ncia prematura. A converg?ncia prematura ocorre quando um enxame (ou uma popula??o) perde a sua capacidade de gerar diversidade e, como consequ?ncia, converge para uma solu??o sub-?tima, prematuramente. Na literatura, existem diversas abordagens que se prop?em a resolver esse problema. Esta tese prop?e uma nova abordagem que ? baseada em um conceito chamado serendipidade que, normalmente, ? aplicado no dom?nio dos Sistemas de Recomenda??o. Para avaliar a viabilidade da adapta??o desse conceito ao novo contexto, uma variante chamada Serendipity-Based Particle Swarm Optimization (SBPSO) foi implementada e, posteriormente, comparada com a Particle Swarm Optimization (PSO) padr?o e algumas variantes apresentadas na literatura. Para realizar os diversos experimentos computacionais, foram utilizadas 16 fun??es de benchmark bastante comuns. Em todos os experimentos, os resultados da SBPSO se mostraram promissores e apresentaram um bom comportamento de converg?ncia, superando a PSO padr?o e as variantes estudadas no que diz respeito ? qualidade da solu??o, ? capacidade de encontrar o ?timo global, ? estabilidade das solu??es e ? capacidade de reiniciar o movimento do enxame ap?s a estagna??o ter sido detectada. / IN the literature, it is common to find many engineering problems which are used to present the effectiveness of the optimization algorithms. Several methods of Bio-Inspired Computing have been proposed as a solution in different contexts of engineering problems. Among these methods, there is a class of algorithms known as Swarm Intelligence. Despite the relative success, most of these algorithms faces a common problem known as premature convergence. It occurs when a swarm loses its ability to generate diversity and consequently converges to a suboptimal solution prematurely. There are several approaches proposed to solve this problem. This doctoral thesis proposes a new approach based on a concept called serendipity. It is usually applied in the field of Recommender Systems. To validate the feasibility of adapting this concept to the new context, a variant called Serendipity-Based Particle Swarm Optimization (SBPSO) has been implemented considering two dimensions of serendipity: chance and sagacity. To evaluate the presented proposal, two sets of computer experiments were performed. Sixteen reference functions which are common in the evaluation of optimization algorithms were used. In the first set of experiments, four functions were used to compare SBPSO to Particle Swarmoptimization (PSO) and some literature variants. In the second ones, twelve other functions were used, but for high dimensionality and a larger number of evaluations of the objective function. In all experiments, the results of the SBPSO were promising and presented a good convergence behaviour with regard to: a) quality of the solution, b) ability to find the global optimum, c) stability of solutions and d) ability to resume the swarmmovement after stagnation has been detected.

Dissertação (mestrado)—Universidade de Brasília, Faculdade de Tecnologia, Departamento de Engenharia Elétrica, 2014. / Submitted by Ana Cristina Barbosa da Silva (annabds@hotmail.com) on 2014-11-25T17:23:31Z No. of bitstreams: 1 2014_RafaelCortesdePaiva.pdf: 7660330 bytes, checksum: eaad53db8e1c76edec638a3e30ee5f3e (MD5) / Approved for entry into archive by Raquel Viana(raquelviana@bce.unb.br) on 2014-11-25T17:58:53Z (GMT) No. of bitstreams: 1 2014_RafaelCortesdePaiva.pdf: 7660330 bytes, checksum: eaad53db8e1c76edec638a3e30ee5f3e (MD5) / Made available in DSpace on 2014-11-25T17:58:54Z (GMT). No. of bitstreams: 1 2014_RafaelCortesdePaiva.pdf: 7660330 bytes, checksum: eaad53db8e1c76edec638a3e30ee5f3e (MD5) / Nesse trabalho foi realizado o estudo de técnicas bio-inspiradas para gerar a marcha de um robô bípede. Foi utilizado o conceito de CPG, Central Pattern Generator (CPG), que é uma rede neural capaz de produzir respostas rítmicas. Elas foram modeladas como osciladores acoplados chamados de osciladores neurais. Para tanto foram utilizados alguns modelos de osciladores, o modelo de Matsuoka, o modelo de Kuramoto e o modelo de Kuramoto com acoplamento entre a dinâmica do oscilador e a dinâmica da marcha. Foram usados dois modelos de robôs, o Bioloid e o NAO. Para otimizar os parâmetros dos osciladores foram utilizados o Algoritmo Genético (AG), o Particle Swarm Optimization (PSO) e o Nondominated sorting Genetic Algorithm II (NSGA-II). Foi utilizada uma função de custo que através de determinadas condições tem como objetivo obter uma marcha eficiente. No NSGA-II, além dessa função de custo, foi utilizada outra função de custo que considera o trabalho realizado pelo robô. Além disso, também foi utilizada a aprendizagem por reforço para treinar um controlador que corrige a postura do robô durante a marcha. Foi possível propor um framework para obter os parâmetros dos osciladores e através dele obter uma marcha estável em ambas as plataformas. Também foi possível propor um framework utilizando aprendizagem por reforço para treinar um controlador para corrigir a postura do robô com a marcha sendo gerado pelo oscilador de Kuramoto com acoplamento. O objetivo do algoritmo foi minimizar a velocidade do ângulo de arfagem do corpo do robô, dessa forma, a variação do ângulo de arfagem também foi minimizada consequentemente. Além disso, o robô andou mais “cautelosamente” para poder manter a postura e dessa forma percorreu uma distância menor do que se estivesse sem o controlador. ______________________________________________________________________________ ABSTRACT / This document describes computational optimized bipedal robot gait generators. Thegaits are applied by a neural oscillator, composed of coupled central pattern generators(CPG), which are neural networks capable of producing rhythmic output. The models ofthe oscillators used were the Matsuoka model, Kuramoto model and Kura moto model withcoupling between the dynamics of the oscillator and dynamics of the gait. Two bipedalrobots, a NAO and a Bioloid, were used. The neural oscillators were optimized with threealgorithms, a Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Nondominatedsorting Genetic Algorithm II (NSGA-II). It was used a fitness function that has theobjective to obtain an efficient gait through some conditions. In NSGA-II, besides this fitnessfunction, another one was used that has the objective to minimize the work done by therobot. Additionally, reinforcement learning techniques were used to train a controller thatcorrects the robots gait posture. It was proposed a framework to obtain the parameters of theoscillators used and obtain efficient gaits in both robots. Also, it was proposed a frameworkusing reinforcement learning to train a controller to correct the robots gait posture. The objective of the algorithm was to minimize the pitch angular velocity, consequently the pitchangle standard deviation was minimized. Additionally, the robot moved with more “caution” and walked less compared with the walk without the posture controller.

Submitted by Almir Azevedo (barbio1313@gmail.com) on 2014-01-15T12:48:16Z No. of bitstreams: 1 dissertacao_mestrado_ien_2008_04.pdf: 1317159 bytes, checksum: c510f22d0bfa406fdceeb4cdbb80e43f (MD5) / Made available in DSpace on 2014-01-15T12:48:16Z (GMT). No. of bitstreams: 1 dissertacao_mestrado_ien_2008_04.pdf: 1317159 bytes, checksum: c510f22d0bfa406fdceeb4cdbb80e43f (MD5) Previous issue date: 2008 / O projeto de modelos em escala reduzida tem sido empregada por engenheiros de vários setores como indústria naval, indústria aeroespacial, petrolífera, indústria nuclear e outras. Modelos em escala reduzida são usados em experimentos porque são economicamente mais atraentes do que seus próprios protótipos (escala real), e em muitos casos também são mais baratos e, na maioria das vezes, mais fáceis de serem construídos fornecendo uma maneira de se conduzir o projeto em escala real permitindo investigações e análises indiretas no sistema em escala real. Um modelo em escala reduzida (ou experimento) deve ser capaz de representar todos os fenômenos físicos que ocorrem e ocorrerão no sistema real em condições de operação, neste caso o modelo em escala reduzida é dito similar. Existem alguns métodos para se projetar um modelo em escala reduzida, e destes, dois métodos são básicos : o método empírico que é baseado na habilidade do profissional especialista para determinar quais são as grandezas físicas relevantes para o modelo desejado, e o método das equações diferenciais que é baseado na descrição matemática do protótipo (ou experimento em escala real) para o modelo. Aplicando uma técnica matemática à equação ou equações diferenciais que descrevem o comportamento do protótipo a partir de leis físicas e assim ressaltando as grandezas físicas (quantidades) relevantes para o problema do projeto do modelo em escala reduzida, e assim o problema pode ser tratado como um problema de otimização. Muitas técnicas de otimização como Algoritmo Genético, por exemplo, tem sido desenvolvidas para solucionar esta classe de problemas e tem também sido aplicadas ao projeto do modelo em escala reduzida. Neste trabalho, é realizada a investigação do uso da técnica de otimização por enxame de partículas, como ferramenta (alternativa) de otimização, no projeto termohidráulico do núcleo de reator PWR em escala reduzida, em regime de circulação forçada e condições normais de operação. Uma comparação de desempenho entre as técnicas GA e PSO é realizada assim como uma comparação entre seus resultados. Os resultados obtidos mostram que a técnica de otimização investigada é uma ferramenta promissora para o projeto de experimentos ou equipamentos em escala reduzida, apresentando vantagens sobre outras técnicas. / The reduced scale models design have been employed by engineers from several different industries fields such as offshore, spatial, oil extraction, nuclear industries and others. Reduced scale models are used in experiments because they are economically attractive than it’s own prototype (real scale) because in many cases they are cheaper than a real scale one and most of time they are also easier to build providing a way to lead the real scale design allowing indirect investigations and analysis to the real scale system (prototype). A reduced scale model (or experiment) must be able to represent all physical phenomena that occurs and further will do in the real scale one under operational conditions, e.g., in this case the reduced scale model is called similar. There are some different methods to design a reduced scale model and from those two are basic : the empiric method based on the expert’s skill to determine which physical measures are relevant to the desired model; and the differential equation method that is based on a mathematical description of the prototype (real scale system) to model. Applying a mathematical technique to the differential equation that describes the prototype then highlighting the relevant physical measures so the reduced scale model design problem may be treated as an optimization problem. Many optimization techniques as Genetic Algorithm (GA), for example, have been developed to solve this class of problems and have also been applied to the reduced scale model design problem as well. In this work, Particle Swarm Optimization (PSO) technique is investigated as an alternative optimization tool for such problem. In this investigation a computational approach, based on particle swarm optimization technique (PSO), is used to perform a reduced scale two loop Pressurized Water Reactor (PWR) core, considering 100% of nominal power operation on a forced flow cooling circulation and non-accidental operating conditions. A performance comparison between GA and PSO techniques is performed as it’s obtained results to this problem. Obtained results shows that the proposed optimization technique (PSO) is a promising tool for a reduced scale experiments or equipments design, presenting advantages over other techniques.

Context. Building autonomous racing car controllers is a growing field of computer science which has been receiving great attention lately. An approach named Artificial Potential Fields (APF) is used widely as a path finding and obstacle avoidance approach in robotics and vehicle motion controlling systems. The use of APF results in a collision free path, it can also be used to achieve other goals such as overtaking and maneuverability. Objectives. The aim of this thesis is to build an autonomous racing car controller that can achieve good performance in terms of speed, time, and damage level. To fulfill our aim we need to achieve optimality in the controller choices because racing requires the highest possible performance. Also, we need to build the controller using algorithms that does not result in high computational overhead. Methods. We used Particle Swarm Optimization (PSO) in combination with APF to achieve optimal car controlling. The Open Racing Car Simulator (TORCS) was used as a testbed for the proposed controller, we have conducted two experiments with different configuration each time to test the performance of our APF- PSO controller. Results. The obtained results showed that using the APF-PSO controller resulted in good performance compared to top performing controllers. Also, the results showed that the use of PSO proved to enhance the performance compared to using APF only. High performance has been proven in the solo driving and in racing competitions, with the exception of an increased level of damage, however, the level of damage was not very high and did not result in a controller shut down. Conclusions. Based on the obtained results we have concluded that the use of PSO with APF results in high performance while taking low computational cost.