On the Type II Equivalency of Bio-inspired Swarm Type I Fuzzy Systems

Type-۲ fuzzy systems aim to build higher robustness in decision making by providing a mechanism to manage the vagueness and imprecision of information, a step further in the direction of better representation of information than the more traditional type-۱ fuzzy systems in which precisiation of information was the ultimate aim. Reportedly, this property has led to the increased robustness of type-۲ fuzzy decision-making systems, while also maintaining or exceeding the high performance of their type-۱ fuzzy predecessors. But type-۲ fuzzy systems are haunted by high computational cost, rendering them incapable of reaching timely decisions in real time systems without considerable simplification. In this paper, we propose that a coordinated swarm of many type-۱ fuzzy systems can approximate a type-۲ fuzzy system. Hence, their computational order of complexity is reduced to that of type I fuzzy systems. And if there is a large enough number of such simple type-۱ fuzzy systems, their aggregate is equivalent to that of a type-۲ fuzzy system. For this purpose, we use a wavy-slice representation of type-۲ fuzzy systems with a decision criterion and take biological inspiration from the factors combinatorial mechanism in genome structure. Simulation studies indicate the merits of the proposed method in the classic truck backer-upper benchmark problem.

کلیدواژه ها:

Genome Structure ، Type-۲ Fuzzy Systems ، Wavy-slice Representation ، Swarm Intelligence