Application of the genetic algorithms to the synthesis of the uniform induction gradient in MRI magnets

The advantages and the robustness of genetic algorithms in the solution of a specific optimal design problem have been investigated. The algorithm implemented has been tested to solve the inverse problem of the synthesis of the magnetic configuration for magnetic resonance imaging; the required configuration consists of a uniform induction field gradient in the bi-dimensional domain located between a couple of symmetric solenoids. The sensitivity analysis of the obtained results has been also performed.