Search

Convolutional Neural Networks (CNN) are novel techniques with significant performance in object detection and classification. An open research problem on CNN is the automatic finding of adequate architectures, which is usually done by hand. Metaheuristic algorithms are techniques that find optimal solutions in heuristic manner to problems where the knowledge is limited or almost nonexistent, such as finding optimal CNN architectures. In this chapter, we propose a framework that utilizes the micro genetic algorithm to find CNN architectures in the shortest possible time. The proposal is tested over three simple study cases known by the research community (MNIST, MNIST-Fashion, and MNIST-RB), and compared against two different frameworks from the literature: psoCNN, and simple genetic algorithm. The results show a better performance of the architectures found by our framework in terms of accuracy and processing time.

Pipe design is one of the most significant research lines in the area of parabolic semi-cylindrical solar collectors. The main idea behind pipe design is to increase the capillarity angle by expanding the total area being heated, therefore boosting the work capacity of the device. Such capillarity depends on several factors, whose numerical calculations are highly complex. Moreover, some of those variables are integers, whereas some others are real; hence, it is necessary to use optimization techniques that are capable of searching in those numerical spaces. There are several optimization tools that allow individual codification as binary strings, granting the coding of integer, real, or any other, as part of the same individual. Consequently, in this paper we propose the comparison of four metaheuristics when they are utilized to maximize the capillarity angle of the pipe in a parabolic trough. Experimental results show a better performance of binary particle swarm optimization when compared against the other techniques, achieving improvements in the capillarity angle of on average 11 % in comparison with a similar study.

Clustering is one of the most popular methods of machine learning. The process of clustering involves the division of a set of abstract objects into a certain number of groups which integrated with objects of similar characteristics. Therefore, a cluster integrates objects which are similar to them, but dissimilar to the elements that belong to the rest of the clusters. Several clustering methods have proposed in the literature with different performance levels. All these techniques use as similarity criterion the Euclidean distance among cluster elements. However, there exist diverse scenarios where the Euclidean distance cannot be utilized appropriately to separate the elements in groups. Under such conditions, traditional cluster methods cannot directly apply. On the other hand, the operations of dilate and erode are a set of non-linear operators that modify the shape of a data group in the feature space, to obtain a monolithic object. Although morphological operations have demonstrated its importance in several engineering fields as image processing, its use as a clustering technique has been practically overlooked. In this work, an alternative clustering algorithm is proposed to group elements without considering the distance as a similarity criterion. In our approach, the data were separated into different groups by considering morphological operations. Under this scheme, the procedure allows the integration of data points, which present a spatial connection. Since the proposed algorithm does not use the distance in its functioning, it solves complex clustering problems which traditional clustering algorithms cannot.

Photovoltaic systems, wind turbines and battery storage systems are all commonly present in energy self-sufficiency and independent producers schemes, since they support distributed generation designs. For both cases, single phase inverters provide power conditioning in order to inject micro-sources power to the grid, demanding multiple tasks such as maximum power tracking, DC to AC voltage conversion, electrical signal filtering and network synchronization, among others. The massive integration of such devices demands an assessment on their impact over the low voltage distribution network to which they are connected to. Hence, an accurate and efficient mathematical modeling is required, for both steady and transient state, in order to provide a robust dynamical simulation to support the designing of coordinated protection schemes and operational control algorithms. The analysis also contributes to ensure network stability within the required quality issues. Therefore, this paper extends the dynamic phasors technique that is a widely employed method for modeling oscillatory systems, in order to develop, simulate and analyze the mathematical model of single phase full bridge inverter. A comparison between different types of numeric methods is made, to find the better option depending of the finality of the study. The technique focuses on three frequencies of interest: the network rate, the boost and the inverter stage frequencies. Based on dynamic phasors information, two PI controls are designed to control the DC an AC voltages, and new formulas for calculation of DC and AC powers are presented. Simulation results in Matlab and Simulink demonstrate the effectiveness of the work.

in digital images is considered an important and complex task for the computer vision community. Consequently, a tremendous amount of research has been devoted to find an optimal circle detector. This article presents an algorithm for the automatic detection of circular shapes from complicated and noisy images with no consideration of the conventional Hough transform principles. The proposed algorithm is based on newly developed Artificial Immune Optimisation (AIO) technique, known as the Clonal Selection Algorithm (CSA). The CSA is an effective method for searching and optimising following the Clonal Selection Principle (CSP) in the human immune system which generates a response according to the relationship between antigens (Ags), i.e. patterns to be recognised and antibodies (Abs), i.e. possible solutions. The algorithm uses the encoding of three points as candidate circles (x,y,r) over the edge image. An objective function evaluates if such candidate circles (Ab) are actually present in the edge image (Ag). Guided by the values of this objective function, the set of encoded candidate circles are evolved using the CSA so that they can fit to the actual circles on the edge map of the image. Experimental results over several synthetic as well as natural images with varying range of complexity validate the efficiency of the proposed technique with regard to accuracy, speed and robustness.

In several machine vision problems, a relevant issue is the estimation of homographies between two different perspectives that hold an extensive set of abnormal data. A method to find such estimation is the random sampling consensus (RANSAC); in this, the goal is to maximize the number of matching points given a permissible error (Pe), according to a candidate model. However, those objectives are in conflict: a low Pe value increases the accuracy of the model but degrades its generalization ability that refers to the number of matching points that tolerate noisy data, whereas a high Pe value improves the noise tolerance of the model but adversely drives the process to false detections. This work considers the estimation process as a multiobjective optimization problem that seeks to maximize the number of matching points whereas Pe is simultaneously minimized. In order to solve the multiobjective formulation, two different evolutionary algorithms have been explored: the Nondominated Sorting Genetic Algorithm II (NSGA-II) and the Nondominated Sorting Differential Evolution (NSDE). Results considering acknowledged quality measures among original and transformed images over a well-known image benchmark show superior performance of the proposal than Random Sample Consensus algorithm. © 2016 Valentín Osuna-Enciso et al.

In image analysis, segmentation is considered one of the most important steps. Segmentation by searching threshold values assumes that objects in a digital image can be modeled through distinct gray level distributions. In this chapter it is proposed the use of a bio-inspired algorithm, called Allostatic Optimisation (AO), to solve the multi threshold segmentation problem. Our approach considers that an histogram can be approximated by a mixture of Cauchy functions, whose parameters are evolved by AO. The contributions of this chapter are on three fronts, by using: a Cauchy mixture to model the original histogram of digital images, the Hellinger distance as an objective function, and AO algorithm. In order to illustrate the proficiency and robustness of the proposed approach, it has been compared to the well-known Otsu method, over several standard benchmark images.

In the renewable energy generation, several processes require the integration of a set of advanced techniques in order to find optimal solutions. Dynamic estimation, stabilizing control for disturbance rejection, optimization for control effort, and parameter tuning are techniques used to address the whole process requirements and obtain optimal results. In this paper, an optimal control strategy for a maximum biofuel production in the presence of disturbances is proposed. First, an integrated optimal control strategy to maximize biofuel production in the presence of disturbances is proposed. Second, due to its high nonlinearity, complex nature, and multiplicity of equilibrium points, a biological process for biofuel generation is described in order to demonstrate the efficiency of the optimal control strategy. A nonlinear discrete-time neural observer for unknown nonlinear systems in the presence of external disturbances and parameter uncertainties is used to estimate unmeasurable variables. An inverse o...

Hough transform (HT) has been the most common method for circle detection exhibiting robustness but adversely demanding a considerable computational load and large storage. Alternative approaches include heuristic methods that employ iterative optimization procedures for detecting multiple circles under the inconvenience that only one circle can be marked at each optimization cycle demanding a longer execution time. On the other hand, Learning Automata (LA) is a heuristic method to solve complex multi-modal optimization problems. Although LA converges to just one global minimum, the final probability distribution holds valuable information regarding other local minima which have emerged during the optimization process. The detection process is considered as a multi-modal optimization problem, allowing the detection of multiple circular shapes through only one optimization procedure. The algorithm uses a combination of three edge points as parameters to determine circles candidates. A reinforcement signal determines if such circle candidates are actually present at the image. Guided by the values of such reinforcement signal, the set of encoded candidate circles are evolved using the LA so that they can fit into actual circular shapes over the edge-only map of the image. The overall approach is a fast multiple-circle detector despite facing complicated conditions., Comment: 30 Pages. arXiv admin note: text overlap with arXiv:1405.5406

Energy systems with renewable sources are used around the world in order to satisfy both off-grid and on-grid load demands, and are commonly coupled to conventional sources. A good behavior of this kind of systems depends on the renewable sources availability that includes the solar irradiance and the wind speed, as well as the profile variations over the energy demand. Their main objective is to satisfy the load demand while minimizing the use of conventional sources, reducing pollutant emissions and storing the energy excess for deficit conditions. This paper presents modeling, neural forecasting and optimal sizing for hybrid energy systems, which are proposed to minimize both the overall annual cost and the use of conventional sources, which in turn represents reduction of pollutant emissions. In this paper, the use of renewable sources along with load demand variations are predicted by a High Order Neural Network trained with an Extended Kalman Filter, whereas the optimal sizing is calculated by using both a Clonal Selection Algorithm and a Genetic Algorithm. The efficiency of using neural forecasting data is illustrated through a simulation with the results showing the effectiveness of both optimization algorithms for calculating an optimal sizing of the hybrid system, which ultimately represents an optimal cost-effective system.

Bio-inspired computing has lately demonstrated its usefulness with remarkable contributions to shape detection, optimization, and classification in pattern recognition. Similarly, multithreshold selection has become a critical step for image analysis and computer vision sparking considerable efforts to design an optimal multi-threshold estimator. This paper presents an algorithm for multi-threshold segmentation which is based on the artificial immune systems(AIS) technique, also known as theclonal selection algorithm (CSA). It follows the clonal selection principle (CSP) from the human immune system which basically generates a response according to the relationship between antigens (Ag), that is, patterns to be recognized and antibodies (Ab), that is, possible solutions. In our approach, the 1D histogram of one image is approximated through a Gaussian mixture model whose parameters are calculated through CSA. Each Gaussian function represents a pixel class and therefore a thresholding point. Unlike the expectation-maximization (EM) algorithm, the CSA-based method shows a fast convergence and a low sensitivity to initial conditions. Remarkably, it also improves complex time-consuming computations commonly required by gradient-based methods. Experimental evidence demonstrates a successful automatic multi-threshold selection based on CSA, comparing its performance to the aforementioned well-known algorithms.

La computación basada en principios físicos recientemente ha ganado respeto en la comunidad científi ca. Esta área emergente, en poco tiempo ha logrado desarrollar un amplio rango de técnicas y métodos que han servido pararesolver diversos problemas, considerados como complejos. Por otra parte, la detección automática de círculos en imágenes se considera una tarea importante, es por esto que se han realizado un gran número de trabajos tratando de encontrar el detector de círculos óptimo. Este artículo presenta un nuevo algoritmo para la detección de primitivas circulares contenidas en imágenes sin la consideración de la transformada de Hough. El algoritmo propuesto está basado en un nuevo enfoque inspirado en principios físicos llamado, Electromagnetism-Like Optimization (EMO), el cual es un método heurístico que emplea algunos principios de la teoría del electromagnetismo para resolver problemas complejos de optimización. En el algoritmo EMO las soluciones se construyen considerando la atracción y repulsión electromagnética entre las partículas cargadas; dicha carga representa la afi nidad que tiene cada partícula con la solución. El algoritmo de detección de círculos emplea una codifi cación de tres puntos no colineales, dichos puntos representan los círculos candidatos sobre una imagen que sólo contiene sus bordes. Empleando una función objetivo, el conjunto de círculos candidatos considerados como partículas cargadas, son operados por medio del algoritmo EMO hasta que logren coincidir con los círculos existentes en la imagen real. Los resultados experimentales en diversas imágenes complejas validaron la efi ciencia de la técnica propuesta en cuanto a su exactitud, velocidad y robustez.

Los glóbulos blancos (GB) o leucocitos juegan un papel importante en el diagnóstico de distintas enfermedades. Aunque técnicas de proce- samiento digital de imágenes han ayudado parcialmente al análisis de tales células, prevalecen distintas complicaciones en la detección de GB, debido a variaciones en forma, tamaño, así como iluminaciones no adecuadas inherentes a la preparación del frotis sanguíneo. Por otra parte, los sistemas inmunes artificiales (SIA), basados en la manera en que el sistema inmunológico natural optimiza sus funciones para la detección de antígenos, han sido aplicados con éxito en la solu- ción de distintos problemas de optimización, produciendo resultados superiores comparados con las técnicas clásicas. El enfoque más usado de SIA es el algoritmo de selección clonal (ASC), el cual permite proliferar a aquellas soluciones (anticuerpos) que mejor resuelven a una determinada función de desempeño (antígeno). Este artículo presenta un algoritmo para la segmentación, detección y medición de leucocitos en imágenes de frotis sanguíneo. El algoritmo usa dos sistemas ASC, uno para segmentación y el otro para detección de leucocitos. El problema de detección es considerado en este traba- jo como un problema de optimización entre el leucocito y la forma circular que mejor lo aproxime. El algoritmo utiliza la codificación de tres puntos de la imagen de bordes para modelar los leucocitos can- didatos. Una función de costo evalúa si tales leucocitos candidatos están realmente en la imagen. Siguiendo los valores de tal función de costo, el grupo de leucocitos candidatos es modificado usando ASC hasta aproximarse a los leucocitos reales presentes en la imagen de bordes. Los resultados obtenidos en comparación a otros algoritmos usados para la misma tarea validan la eficiencia de esta propuesta en exactitud, velocidad y robustez.

Over the last decade, several bio-inspired algorithms have emerged for solving complex optimisation problems. Since the performance of these algorithms present a suboptimal behaviour, a tremendous amount of research has been devoted to find new and better optimisation methods. On the other hand, allostasis is a medical term recently coined which explains how the configuration of the internal state IS in different organs allows reaching stability when an unbalance condition is presented. In this paper, a novel biologically-inspired algorithm called allostatic optimisation AO is proposed for solving optimisation problems. In AO, individuals emulate the IS of different organs. In the approach, each individual is improved by using numerical operators based on the biological principles of the allostasis mechanism. The proposed method has been compared to other well-known optimisation algorithms. The results show good performance of the proposed method when searching for a global optimum of several benchmark functions.