Search

Mammalian immune system and especially clonal selection principle, responsible for coping with external intruders, is an inspiration for a set of heuristic optimization algorithms. In this paper we focus our attention on an instance of a clonal selection algorithm called BCA. This algorithm admits very good exploratory abilities when solving stationary optimization problems. We try to explain this capability studying the behavior of the mutation operator being distinguishable feature of BCA. Later we apply it to solve non-stationary optimization problem.

In order to solve a Non-Stationary Optimization Problem (NSOP) it is necessary that the used algorithms have a set of suitable properties for being able to dynamically adapt the search to the changing fitness landscape. Our aim in this work is to improve our knowledge of existing canonical algorithms (steady-state, generational, and structured –cellular– genetic algorithms) in such a scenario. We study the behavior of these algorithms in a basic Dynamic Knapsack Problem, and utilize quantitative metrics for analyzing the results. In this work, we analyze the role of the mutation operator in the three algorithms and the impact of the frequency of dynamic changes in the resulting difficulty of the problem. Our conclusions outline that the steady-state GA is the best in fast adapting its search to a new problem definition, while the cellular GA is the best in preserving diversity to finally get accurate solutions. The generational GA is a tradeoff algorithm showing performances in between the other two.

This paper is devoted to advanced mutation operators for C# source code. They deal with object-oriented (OO mutations) and other complex features of the code. They require structural information about a code, unlike the standard mutations. Applicability of OO operators in C# is compared with those for other OO languages. Operators for specific features of C# language are also proposed. The detailed specification of operators can be provided in terms of pre- and post-conditions of a program transformation. Based on the operators’ specification, the generation of mutated C# programs can be automated.

This paper proposes a novel Genetic Algorithms (GAs) approach for a near-optimal path planning of a mobile robot in a greenhouse. The chromosome encoding features in inverse proportion between research spaces of GAs and complexity of obstacles. The fitness evaluation is designed for both incomplete and complete paths to guide the evolutional direction. The crossover and mutation operators are trimmed to he path planning problem. Two operators were presented to promote the effectiveness of evolution of problem-specific GAs. The simulation results obtained is satisfactory.

The Petzval objective was the first photographic objective to be deliberately designed rather than being put together by an empirical selection of lenses out of box. As originally designed by Joseph Petzval it took the form of two doublets widely apart with a central aperture diaphragm. The basic form of the Petzval objective has been retained in many present-day applications, but detailed changes have been made in the doublets. In some instances, for example, the doublets have been replaced by doublets plus associated singlets. In other variants of the basic form a third doublet has been added.

Mutation analysis on model checking specifications is a recent development. This approach mutates a specification, then applies a model checker to compare the mutants with the original specification to automatically generate tests or evaluate coverage. The properties of specification mutation operators have not been explored in depth. We report our work on theoretical and empirical comparison of these operators. Our future plans include studying how the form of a specification influences the results, finding relations between different operators, and validating the method against independent metrics.

Verification, validation and testing techniques developed for use with conventional development practices, are not always applicable when developing knowledge-based software. This paper presents an experimental framework to determine whether a technique is applicable or not, based on concepts from mutation testing. The framework itself comprises of a number of steps guiding the researcher/practitioner in the assessment process. Mutation testing is used to simulate faults in an example programme to determine the technique’s ability to detect them. The framework has been applied to two techniques: control-flow analysis and cause-effect graphing. The conclusion is that the framework gives a good basis for a qualitative assessment of the applicability and efficiency of applying specific traditional VV&T techniques to knowledge-based components.

In Chapters 5 and 6 “classical” methods for BDD minimization that are supported by most available packages and recent refinements of these methods have been presented.

This paper presents some recent results that were obtained when a basic genetic algorithm (GA) for optimizing the cost of electrical wire harnesses was modified. These modifications included the incorporation of two operators that were specific for the problem being solved: a) a gauge propagation operator, and b) an operator that attempts to improve a solution by randomly changing wire gauges associated with a particular device of the harness. In addition, the modified GA included the implementation of a meta-architecture that was useful to overcome the problem of finding a set of good input parameters for running the single-layered GA. These modifications differ from other general purpose techniques that have been suggested for improving the search in GAs. Results obtained with the modified GA for an example harness showed that the modifications were helpful for improving the effectiveness and efficiency of the basic GA.

In this paper we use simulated evolution to solve three design theoretic problems, direct decomposition of a complete graph into smaller complete graphs, decomposition of a complete graph into smaller complete graphs with difference sets, and the seating of twelve people at three four person tables for multiple nights in an even handed fashion. The paper is intended as a tutorial for someone familiar with design theory but unfamiliar with genetic algorithms.