blems of Synthesis of Adaptive Filter Techniques of Genetic Algorithms to Software in C++ OS MSVS

Abstract

. In solving the problem of radiation stable signals with constant spectral composition is required to synthesize the appropriate signal model equation. For the synthesis is proposed to use one of the types of genetic algorithms — a method of group account of arguments (GMDH). In fact, synthesized by this method the equation is the optimal filter — a private performance of the series expansion Lotka–Volterra. The results obtained expansion — the result of a genetic algorithm. The problems in the synthesis of the optimal filter are usually separate data, usually generate models of applicants, the method of obtaining the coefficients of the model and the form of the quadratic model selection criterion. The report provides algorithms for solving these problems, implemented in C++ and runs under OS MSVS. Of particular interest is the modified algorithm for determining the coefficients Seidel optimal model capable of oper-ating with a poorly-defined matrices having determinant value close or equal to zero.