Detecção de falhas em estruturas inteligentes usando otimização por nuvem de partículas: fundamentos e estudo de casos

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

What is the importance of selecting features for non-technical losses identification?

Although non-technical losses automatic identification has been massively studied, the problem of selecting the most representative features in order to boost the identification accuracy has not attracted much attention in this context. In this paper, we focus on this problem applying a novel feature selection algorithm based on Particle Swarm Optimization and Optimum-Path Forest. The results demonstrated that this method can improve the classification accuracy of possible frauds up to 49% in some datasets composed by industrial and commercial profiles. © 2011 IEEE.

Kappa-PSO-ARTMAP Fuzzy: uma metodologia para detecção de intrusos baseado em seleção de atributos e otimização de parâmetros numa rede neural ARTMAP Fuzzy

Pós-graduação em Engenharia Elétrica - FEIS

Desenvolvimento de metodologias para a localização de intruso em ambientes indoor

O presente trabalho propõe metodologias para detectar a presença e localizar um intruso em ambientes indoor, 2-D e 3-D, sendo que neste último, utiliza-se um sistema cooperativo de antenas e, em ambos os casos, o sistema é baseado em radares multiestáticos. Para obter uma alta resolução, o radar opera com pulsos UWB, que possuem amplitude espectral máxima em 1 GHz para ambientes 2-D e, pulsos de banda larga com frequências entre 200 MHz e 500 MHz para ambientes 3-D. A estimativa de localização, para os ambientes bidimensionais, é feita pela técnica de otimização Enxame de Partículas - PSO (Particle Swarm Optimization), pelo método de Newton com eliminação de Gauss e pelo método dos mínimos quadrados com eliminação de Gauss. Para o ambiente tridimensional, foi desenvolvida uma metodologia vetorial que estima uma possível região de localização do intruso. Para a simulação das ondas eletromagnéticas se utiliza o método numérico FDTD (Diferenças Finitas no Domínio do Tempo) associado à técnica de absorção UPML (Uniaxial Perfectly Matched Layer) com o objetivo de truncar o domínio de análise simulando uma propagação ao infinito. Para a análise do ambiente em 2-D foi desenvolvido o ACOR-UWB-2-D e para o ambiente 3-D foi utilizado o software LANE SAGS.

Imputação de dados baseado em otimização por enxame de partículas considerando os principais mecanismos de ausência de dados

Durante o processo de extração do conhecimento em bases de dados, alguns problemas podem ser encontrados como por exemplo, a ausência de determinada instância de um atributo. A ocorrência de tal problemática pode causar efeitos danosos nos resultados finais do processo, pois afeta diretamente a qualidade dos dados a ser submetido a um algoritmo de aprendizado de máquina. Na literatura, diversas propostas são apresentadas a fim de contornar tal dano, dentre eles está a de imputação de dados, a qual estima um valor plausível para substituir o ausente. Seguindo essa área de solução para o problema de valores ausentes, diversos trabalhos foram analisados e algumas observações foram realizadas como, a pouca utilização de bases sintéticas que simulem os principais mecanismos de ausência de dados e uma recente tendência a utilização de algoritmos bio-inspirados como tratamento do problema. Com base nesse cenário, esta dissertação apresenta um método de imputação de dados baseado em otimização por enxame de partículas, pouco explorado na área, e o aplica para o tratamento de bases sinteticamente geradas, as quais consideram os principais mecanismos de ausência de dados, MAR, MCAR e NMAR. Os resultados obtidos ao comprar diferentes configurações do método à outros dois conhecidos na área (KNNImpute e SVMImpute) são promissores para sua utilização na área de tratamento de valores ausentes uma vez que alcançou os melhores valores na maioria dos experimentos realizados.

A influência do controlador gupfc no sistema elétrico de potência visando a estabilidade a pequenas perturbações

Pós-graduação em Engenharia Elétrica - FEIS

Global optimization using a genetic algorithm with hierarchically structured population

This paper applies a genetic algorithm with hierarchically structured population to solve unconstrained optimization problems. The population has individuals distributed in several overlapping clusters, each one with a leader and a variable number of support individuals. The hierarchy establishes that leaders must be fitter than its supporters with the topological organization of the clusters following a tree. Computational tests evaluate different population structures, population sizes and crossover operators for better algorithm performance. A set of known benchmark test problems is solved and the results found are compared with those obtained from other methods described in the literature, namely, two genetic algorithms, a simulated annealing, a differential evolution and a particle swarm optimization. The results indicate that the method employed is capable of achieving better performance than the previous approaches in regard as the two criteria usually employed for comparisons: the number of function evaluations and rate of success. The method also has a superior performance if the number of problems solved is taken into account. (C) 2013 Elsevier B.V. All rights reserved.

Métodos de otimização aplicados no ajuste de ESPS e controladores de amortecimento inseridos no FACTS TCSC em sistemas elétricos de potência

Pós-graduação em Engenharia Elétrica - FEIS

Harmony Search applied for Support Vector Machines Training Optimization

Since the beginning, some pattern recognition techniques have faced the problem of high computational burden for dataset learning. Among the most widely used techniques, we may highlight Support Vector Machines (SVM), which have obtained very promising results for data classification. However, this classifier requires an expensive training phase, which is dominated by a parameter optimization that aims to make SVM less prone to errors over the training set. In this paper, we model the problem of finding such parameters as a metaheuristic-based optimization task, which is performed through Harmony Search (HS) and some of its variants. The experimental results have showen the robustness of HS-based approaches for such task in comparison against with an exhaustive (grid) search, and also a Particle Swarm Optimization-based implementation.

Combining meta-learning and search techniques to select parameters for support vector machines

Support Vector Machines (SVMs) have achieved very good performance on different learning problems. However, the success of SVMs depends on the adequate choice of the values of a number of parameters (e.g., the kernel and regularization parameters). In the current work, we propose the combination of meta-learning and search algorithms to deal with the problem of SVM parameter selection. In this combination, given a new problem to be solved, meta-learning is employed to recommend SVM parameter values based on parameter configurations that have been successfully adopted in previous similar problems. The parameter values returned by meta-learning are then used as initial search points by a search technique, which will further explore the parameter space. In this proposal, we envisioned that the initial solutions provided by meta-learning are located in good regions of the search space (i.e. they are closer to optimum solutions). Hence, the search algorithm would need to evaluate a lower number of candidate solutions when looking for an adequate solution. In this work, we investigate the combination of meta-learning with two search algorithms: Particle Swarm Optimization and Tabu Search. The implemented hybrid algorithms were used to select the values of two SVM parameters in the regression domain. These combinations were compared with the use of the search algorithms without meta-learning. The experimental results on a set of 40 regression problems showed that, on average, the proposed hybrid methods obtained lower error rates when compared to their components applied in isolation.

Analog circuit synthesis performing fast Pareto frontier exploration and analysis through 3D graphs

This paper presents a technique for performing analog design synthesis at circuit level providing feedback to the designer through the exploration of the Pareto frontier. A modified simulated annealing which is able to perform crossover with past anchor points when a local minimum is found which is used as the optimization algorithm on the initial synthesis procedure. After all specifications are met, the algorithm searches for the extreme points of the Pareto frontier in order to obtain a non-exhaustive exploration of the Pareto front. Finally, multi-objective particle swarm optimization is used to spread the results and to find a more accurate frontier. Piecewise linear functions are used as single-objective cost functions to produce a smooth and equal convergence of all measurements to the desired specifications during the composition of the aggregate objective function. To verify the presented technique two circuits were designed, which are: a Miller amplifier with 96 dB Voltage gain, 15.48 MHz unity gain frequency, slew rate of 19.2 V/mu s with a current supply of 385.15 mu A, and a complementary folded cascode with 104.25 dB Voltage gain, 18.15 MHz of unity gain frequency and a slew rate of 13.370 MV/mu s. These circuits were synthesized using a 0.35 mu m technology. The results show that the method provides a fast approach for good solutions using the modified SA and further good Pareto front exploration through its connection to the particle swarm optimization algorithm.

Analyzing diffuse scattering with supercomputers

Two new approaches to quantitatively analyze diffuse diffraction intensities from faulted layer stacking are reported. The parameters of a probability-based growth model are determined with two iterative global optimization methods: a genetic algorithm (GA) and particle swarm optimization (PSO). The results are compared with those from a third global optimization method, a differential evolution (DE) algorithm [Storn & Price (1997). J. Global Optim. 11, 341–359]. The algorithm efficiencies in the early and late stages of iteration are compared. The accuracy of the optimized parameters improves with increasing size of the simulated crystal volume. The wall clock time for computing quite large crystal volumes can be kept within reasonable limits by the parallel calculation of many crystals (clones) generated for each model parameter set on a super- or grid computer. The faulted layer stacking in single crystals of trigonal three-pointedstar- shaped tris(bicylco[2.1.1]hexeno)benzene molecules serves as an example for the numerical computations. Based on numerical values of seven model parameters (reference parameters), nearly noise-free reference intensities of 14 diffuse streaks were simulated from 1280 clones, each consisting of 96 000 layers (reference crystal). The parameters derived from the reference intensities with GA, PSO and DE were compared with the original reference parameters as a function of the simulated total crystal volume. The statistical distribution of structural motifs in the simulated crystals is in good agreement with that in the reference crystal. The results found with the growth model for layer stacking disorder are applicable to other disorder types and modeling techniques, Monte Carlo in particular.

Algoritmos de distribución de cargas de proceso en computación con membranas

Con el surgir de los problemas irresolubles de forma eficiente en tiempo polinomial en base al dato de entrada, surge la Computación Natural como alternativa a la computación clásica. En esta disciplina se trata de o bien utilizar la naturaleza como base de cómputo o bien, simular su comportamiento para obtener mejores soluciones a los problemas que los encontrados por la computación clásica. Dentro de la computación natural, y como una representación a nivel celular, surge la Computación con Membranas. La primera abstracción de las membranas que se encuentran en las células, da como resultado los P sistemas de transición. Estos sistemas, que podrían ser implementados en medios biológicos o electrónicos, son la base de estudio de esta Tesis. En primer lugar, se estudian las implementaciones que se han realizado, con el fin de centrarse en las implementaciones distribuidas, que son las que pueden aprovechar las características intrínsecas de paralelismo y no determinismo. Tras un correcto estudio del estado actual de las distintas etapas que engloban a la evolución del sistema, se concluye con que las distribuciones que buscan un equilibrio entre las dos etapas (aplicación y comunicación), son las que mejores resultados presentan. Para definir estas distribuciones, es necesario definir completamente el sistema, y cada una de las partes que influyen en su transición. Además de los trabajos de otros investigadores, y junto a ellos, se realizan variaciones a los proxies y arquitecturas de distribución, para tener completamente definidos el comportamiento dinámico de los P sistemas. A partir del conocimiento estático –configuración inicial– del P sistema, se pueden realizar distribuciones de membranas en los procesadores de un clúster para obtener buenos tiempos de evolución, con el fin de que la computación del P sistema sea realizada en el menor tiempo posible. Para realizar estas distribuciones, hay que tener presente las arquitecturas –o forma de conexión– de los procesadores del clúster. La existencia de 4 arquitecturas, hace que el proceso de distribución sea dependiente de la arquitectura a utilizar, y por tanto, aunque con significativas semejanzas, los algoritmos de distribución deben ser realizados también 4 veces. Aunque los propulsores de las arquitecturas han estudiado el tiempo óptimo de cada arquitectura, la inexistencia de distribuciones para estas arquitecturas ha llevado a que en esta Tesis se probaran las 4, hasta que sea posible determinar que en la práctica, ocurre lo mismo que en los estudios teóricos. Para realizar la distribución, no existe ningún algoritmo determinista que consiga una distribución que satisfaga las necesidades de la arquitectura para cualquier P sistema. Por ello, debido a la complejidad de dicho problema, se propone el uso de metaheurísticas de Computación Natural. En primer lugar, se propone utilizar Algoritmos Genéticos, ya que es posible realizar alguna distribución, y basada en la premisa de que con la evolución, los individuos mejoran, con la evolución de dichos algoritmos, las distribuciones también mejorarán obteniéndose tiempos cercanos al óptimo teórico. Para las arquitecturas que preservan la topología arbórea del P sistema, han sido necesarias realizar nuevas representaciones, y nuevos algoritmos de cruzamiento y mutación. A partir de un estudio más detallado de las membranas y las comunicaciones entre procesadores, se ha comprobado que los tiempos totales que se han utilizado para la distribución pueden ser mejorados e individualizados para cada membrana. Así, se han probado los mismos algoritmos, obteniendo otras distribuciones que mejoran los tiempos. De igual forma, se han planteado el uso de Optimización por Enjambres de Partículas y Evolución Gramatical con reescritura de gramáticas (variante de Evolución Gramatical que se presenta en esta Tesis), para resolver el mismo cometido, obteniendo otro tipo de distribuciones, y pudiendo realizar una comparativa de las arquitecturas. Por último, el uso de estimadores para el tiempo de aplicación y comunicación, y las variaciones en la topología de árbol de membranas que pueden producirse de forma no determinista con la evolución del P sistema, hace que se deba de monitorizar el mismo, y en caso necesario, realizar redistribuciones de membranas en procesadores, para seguir obteniendo tiempos de evolución razonables. Se explica, cómo, cuándo y dónde se deben realizar estas modificaciones y redistribuciones; y cómo es posible realizar este recálculo. Abstract Natural Computing is becoming a useful alternative to classical computational models since it its able to solve, in an efficient way, hard problems in polynomial time. This discipline is based on biological behaviour of living organisms, using nature as a basis of computation or simulating nature behaviour to obtain better solutions to problems solved by the classical computational models. Membrane Computing is a sub discipline of Natural Computing in which only the cellular representation and behaviour of nature is taken into account. Transition P Systems are the first abstract representation of membranes belonging to cells. These systems, which can be implemented in biological organisms or in electronic devices, are the main topic studied in this thesis. Implementations developed in this field so far have been studied, just to focus on distributed implementations. Such distributions are really important since they can exploit the intrinsic parallelism and non-determinism behaviour of living cells, only membranes in this case study. After a detailed survey of the current state of the art of membranes evolution and proposed algorithms, this work concludes that best results are obtained using an equal assignment of communication and rules application inside the Transition P System architecture. In order to define such optimal distribution, it is necessary to fully define the system, and each one of the elements that influence in its transition. Some changes have been made in the work of other authors: load distribution architectures, proxies definition, etc., in order to completely define the dynamic behaviour of the Transition P System. Starting from the static representation –initial configuration– of the Transition P System, distributions of membranes in several physical processors of a cluster is algorithmically done in order to get a better performance of evolution so that the computational complexity of the Transition P System is done in less time as possible. To build these distributions, the cluster architecture –or connection links– must be considered. The existence of 4 architectures, makes that the process of distribution depends on the chosen architecture, and therefore, although with significant similarities, the distribution algorithms must be implemented 4 times. Authors who proposed such architectures have studied the optimal time of each one. The non existence of membrane distributions for these architectures has led us to implement a dynamic distribution for the 4. Simulations performed in this work fix with the theoretical studies. There is not any deterministic algorithm that gets a distribution that meets the needs of the architecture for any Transition P System. Therefore, due to the complexity of the problem, the use of meta-heuristics of Natural Computing is proposed. First, Genetic Algorithm heuristic is proposed since it is possible to make a distribution based on the premise that along with evolution the individuals improve, and with the improvement of these individuals, also distributions enhance, obtaining complexity times close to theoretical optimum time. For architectures that preserve the tree topology of the Transition P System, it has been necessary to make new representations of individuals and new algorithms of crossover and mutation operations. From a more detailed study of the membranes and the communications among processors, it has been proof that the total time used for the distribution can be improved and individualized for each membrane. Thus, the same algorithms have been tested, obtaining other distributions that improve the complexity time. In the same way, using Particle Swarm Optimization and Grammatical Evolution by rewriting grammars (Grammatical Evolution variant presented in this thesis), to solve the same distribution task. New types of distributions have been obtained, and a comparison of such genetic and particle architectures has been done. Finally, the use of estimators for the time of rules application and communication, and variations in tree topology of membranes that can occur in a non-deterministic way with evolution of the Transition P System, has been done to monitor the system, and if necessary, perform a membrane redistribution on processors to obtain reasonable evolution time. How, when and where to make these changes and redistributions, and how it can perform this recalculation, is explained.

Server power modeling for run-time energy optimization of cloud computing facilities

As advanced Cloud services are becoming mainstream, the contribution of data centers in the overall power consumption of modern cities is growing dramatically. The average consumption of a single data center is equivalent to the energy consumption of 25.000 households. Modeling the power consumption for these infrastructures is crucial to anticipate the effects of aggressive optimization policies, but accurate and fast power modeling is a complex challenge for high-end servers not yet satisfied by analytical approaches. This work proposes an automatic method, based on Multi-Objective Particle Swarm Optimization, for the identification of power models of enterprise servers in Cloud data centers. Our approach, as opposed to previous procedures, does not only consider the workload consolidation for deriving the power model, but also incorporates other non traditional factors like the static power consumption and its dependence with temperature. Our experimental results shows that we reach slightly better models than classical approaches, but simul- taneously simplifying the power model structure and thus the numbers of sensors needed, which is very promising for a short-term energy prediction. This work, validated with real Cloud applications, broadens the possibilities to derive efficient energy saving techniques for Cloud facilities.

Contributions to the Resilience Management in the Internet of Things

El auge del "Internet de las Cosas" (IoT, "Internet of Things") y sus tecnologías asociadas han permitido su aplicación en diversos dominios de la aplicación, entre los que se encuentran la monitorización de ecosistemas forestales, la gestión de catástrofes y emergencias, la domótica, la automatización industrial, los servicios para ciudades inteligentes, la eficiencia energética de edificios, la detección de intrusos, la gestión de desastres y emergencias o la monitorización de señales corporales, entre muchas otras. La desventaja de una red IoT es que una vez desplegada, ésta queda desatendida, es decir queda sujeta, entre otras cosas, a condiciones climáticas cambiantes y expuestas a catástrofes naturales, fallos de software o hardware, o ataques maliciosos de terceros, por lo que se puede considerar que dichas redes son propensas a fallos. El principal requisito de los nodos constituyentes de una red IoT es que estos deben ser capaces de seguir funcionando a pesar de sufrir errores en el propio sistema. La capacidad de la red para recuperarse ante fallos internos y externos inesperados es lo que se conoce actualmente como "Resiliencia" de la red. Por tanto, a la hora de diseñar y desplegar aplicaciones o servicios para IoT, se espera que la red sea tolerante a fallos, que sea auto-configurable, auto-adaptable, auto-optimizable con respecto a nuevas condiciones que puedan aparecer durante su ejecución. Esto lleva al análisis de un problema fundamental en el estudio de las redes IoT, el problema de la "Conectividad". Se dice que una red está conectada si todo par de nodos en la red son capaces de encontrar al menos un camino de comunicación entre ambos. Sin embargo, la red puede desconectarse debido a varias razones, como que se agote la batería, que un nodo sea destruido, etc. Por tanto, se hace necesario gestionar la resiliencia de la red con el objeto de mantener la conectividad entre sus nodos, de tal manera que cada nodo IoT sea capaz de proveer servicios continuos, a otros nodos, a otras redes o, a otros servicios y aplicaciones. En este contexto, el objetivo principal de esta tesis doctoral se centra en el estudio del problema de conectividad IoT, más concretamente en el desarrollo de modelos para el análisis y gestión de la Resiliencia, llevado a la práctica a través de las redes WSN, con el fin de mejorar la capacidad la tolerancia a fallos de los nodos que componen la red. Este reto se aborda teniendo en cuenta dos enfoques distintos, por una parte, a diferencia de otro tipo de redes de dispositivos convencionales, los nodos en una red IoT son propensos a perder la conexión, debido a que se despliegan en entornos aislados, o en entornos con condiciones extremas; por otra parte, los nodos suelen ser recursos con bajas capacidades en términos de procesamiento, almacenamiento y batería, entre otros, por lo que requiere que el diseño de la gestión de su resiliencia sea ligero, distribuido y energéticamente eficiente. En este sentido, esta tesis desarrolla técnicas auto-adaptativas que permiten a una red IoT, desde la perspectiva del control de su topología, ser resiliente ante fallos en sus nodos. Para ello, se utilizan técnicas basadas en lógica difusa y técnicas de control proporcional, integral y derivativa (PID - "proportional-integral-derivative"), con el objeto de mejorar la conectividad de la red, teniendo en cuenta que el consumo de energía debe preservarse tanto como sea posible. De igual manera, se ha tenido en cuenta que el algoritmo de control debe ser distribuido debido a que, en general, los enfoques centralizados no suelen ser factibles a despliegues a gran escala. El presente trabajo de tesis implica varios retos que conciernen a la conectividad de red, entre los que se incluyen: la creación y el análisis de modelos matemáticos que describan la red, una propuesta de sistema de control auto-adaptativo en respuesta a fallos en los nodos, la optimización de los parámetros del sistema de control, la validación mediante una implementación siguiendo un enfoque de ingeniería del software y finalmente la evaluación en una aplicación real. Atendiendo a los retos anteriormente mencionados, el presente trabajo justifica, mediante una análisis matemático, la relación existente entre el "grado de un nodo" (definido como el número de nodos en la vecindad del nodo en cuestión) y la conectividad de la red, y prueba la eficacia de varios tipos de controladores que permiten ajustar la potencia de trasmisión de los nodos de red en respuesta a eventuales fallos, teniendo en cuenta el consumo de energía como parte de los objetivos de control. Así mismo, este trabajo realiza una evaluación y comparación con otros algoritmos representativos; en donde se demuestra que el enfoque desarrollado es más tolerante a fallos aleatorios en los nodos de la red, así como en su eficiencia energética. Adicionalmente, el uso de algoritmos bioinspirados ha permitido la optimización de los parámetros de control de redes dinámicas de gran tamaño. Con respecto a la implementación en un sistema real, se han integrado las propuestas de esta tesis en un modelo de programación OSGi ("Open Services Gateway Initiative") con el objeto de crear un middleware auto-adaptativo que mejore la gestión de la resiliencia, especialmente la reconfiguración en tiempo de ejecución de componentes software cuando se ha producido un fallo. Como conclusión, los resultados de esta tesis doctoral contribuyen a la investigación teórica y, a la aplicación práctica del control resiliente de la topología en redes distribuidas de gran tamaño. Los diseños y algoritmos presentados pueden ser vistos como una prueba novedosa de algunas técnicas para la próxima era de IoT. A continuación, se enuncian de forma resumida las principales contribuciones de esta tesis: (1) Se han analizado matemáticamente propiedades relacionadas con la conectividad de la red. Se estudia, por ejemplo, cómo varía la probabilidad de conexión de la red al modificar el alcance de comunicación de los nodos, así como cuál es el mínimo número de nodos que hay que añadir al sistema desconectado para su re-conexión. (2) Se han propuesto sistemas de control basados en lógica difusa para alcanzar el grado de los nodos deseado, manteniendo la conectividad completa de la red. Se han evaluado diferentes tipos de controladores basados en lógica difusa mediante simulaciones, y los resultados se han comparado con otros algoritmos representativos. (3) Se ha investigado más a fondo, dando un enfoque más simple y aplicable, el sistema de control de doble bucle, y sus parámetros de control se han optimizado empleando algoritmos heurísticos como el método de la entropía cruzada (CE, "Cross Entropy"), la optimización por enjambre de partículas (PSO, "Particle Swarm Optimization"), y la evolución diferencial (DE, "Differential Evolution"). (4) Se han evaluado mediante simulación, la mayoría de los diseños aquí presentados; además, parte de los trabajos se han implementado y validado en una aplicación real combinando técnicas de software auto-adaptativo, como por ejemplo las de una arquitectura orientada a servicios (SOA, "Service-Oriented Architecture"). ABSTRACT The advent of the Internet of Things (IoT) enables a tremendous number of applications, such as forest monitoring, disaster management, home automation, factory automation, smart city, etc. However, various kinds of unexpected disturbances may cause node failure in the IoT, for example battery depletion, software/hardware malfunction issues and malicious attacks. So, it can be considered that the IoT is prone to failure. The ability of the network to recover from unexpected internal and external failures is known as "resilience" of the network. Resilience usually serves as an important non-functional requirement when designing IoT, which can further be broken down into "self-*" properties, such as self-adaptive, self-healing, self-configuring, self-optimization, etc. One of the consequences that node failure brings to the IoT is that some nodes may be disconnected from others, such that they are not capable of providing continuous services for other nodes, networks, and applications. In this sense, the main objective of this dissertation focuses on the IoT connectivity problem. A network is regarded as connected if any pair of different nodes can communicate with each other either directly or via a limited number of intermediate nodes. More specifically, this thesis focuses on the development of models for analysis and management of resilience, implemented through the Wireless Sensor Networks (WSNs), which is a challenging task. On the one hand, unlike other conventional network devices, nodes in the IoT are more likely to be disconnected from each other due to their deployment in a hostile or isolated environment. On the other hand, nodes are resource-constrained in terms of limited processing capability, storage and battery capacity, which requires that the design of the resilience management for IoT has to be lightweight, distributed and energy-efficient. In this context, the thesis presents self-adaptive techniques for IoT, with the aim of making the IoT resilient against node failures from the network topology control point of view. The fuzzy-logic and proportional-integral-derivative (PID) control techniques are leveraged to improve the network connectivity of the IoT in response to node failures, meanwhile taking into consideration that energy consumption must be preserved as much as possible. The control algorithm itself is designed to be distributed, because the centralized approaches are usually not feasible in large scale IoT deployments. The thesis involves various aspects concerning network connectivity, including: creation and analysis of mathematical models describing the network, proposing self-adaptive control systems in response to node failures, control system parameter optimization, implementation using the software engineering approach, and evaluation in a real application. This thesis also justifies the relations between the "node degree" (the number of neighbor(s) of a node) and network connectivity through mathematic analysis, and proves the effectiveness of various types of controllers that can adjust power transmission of the IoT nodes in response to node failures. The controllers also take into consideration the energy consumption as part of the control goals. The evaluation is performed and comparison is made with other representative algorithms. The simulation results show that the proposals in this thesis can tolerate more random node failures and save more energy when compared with those representative algorithms. Additionally, the simulations demonstrate that the use of the bio-inspired algorithms allows optimizing the parameters of the controller. With respect to the implementation in a real system, the programming model called OSGi (Open Service Gateway Initiative) is integrated with the proposals in order to create a self-adaptive middleware, especially reconfiguring the software components at runtime when failures occur. The outcomes of this thesis contribute to theoretic research and practical applications of resilient topology control for large and distributed networks. The presented controller designs and optimization algorithms can be viewed as novel trials of the control and optimization techniques for the coming era of the IoT. The contributions of this thesis can be summarized as follows: (1) Mathematically, the fault-tolerant probability of a large-scale stochastic network is analyzed. It is studied how the probability of network connectivity depends on the communication range of the nodes, and what is the minimum number of neighbors to be added for network re-connection. (2) A fuzzy-logic control system is proposed, which obtains the desired node degree and in turn maintains the network connectivity when it is subject to node failures. There are different types of fuzzy-logic controllers evaluated by simulations, and the results demonstrate the improvement of fault-tolerant capability as compared to some other representative algorithms. (3) A simpler but more applicable approach, the two-loop control system is further investigated, and its control parameters are optimized by using some heuristic algorithms such as Cross Entropy (CE), Particle Swarm Optimization (PSO), and Differential Evolution (DE). (4) Most of the designs are evaluated by means of simulations, but part of the proposals are implemented and tested in a real-world application by combining the self-adaptive software technique and the control algorithms which are presented in this thesis.