Optical computing architecture is an effective basis for the creation of high-performance computing devices with the possibility of widespread use of parallel structures and high-speed optical logic elements, as well as reducing the number of switching components and energy consumption. The authors develop the computational architecture of logical operations based on (a) the representation of fuzzy information as a color quantum, (b) the use of optical transformations of ray radiation as natural operations of disjunction and conjunction, as well as (c) the basis for constructing more complex logical operations for deriving decisions. The concepts of simple and complex decisions, blocking and new decisions, as well as decisions based on the negation of the main output solution have been formed and clarified. Optical structural components for the logical operations of each of the considered decisions, as well as more complex combinations of logical outputs, are formed. A mathematical description of algorithmic inference procedures is given. The use of the basic laws of optical physics and strictly mathematical apparatus in the synthesis of computational structural components is the basis for the practical implementation of the proposed computational architecture.