Identification of optimal polymeric blend for cables using Particle Swarm Optimization method

The individual electrical and mechanical properties of polymeric materials such as Silicone Rubber (SR) and Ethylene Propylene Diene Monomer (EPDM) are found to be unsatisfactory for cable applications. Blending of polymers is the well known method to fully utilize the superior properties of both the materials. Blends of SR/EPDM are prepared with different blend ratios of the constituent polymers. The blends are characterized as per ASTM and IEC standards for electrical properties like volume, surface resistivity, arc resistance and comparative tracking index and mechanical properties like tensile strength and elongation at break. The choice of optimal blend (OB) is actually a tradeoff between the mechanical and electrical properties. Hence, there is a need for intelligence techniques like Particle swarm optimization (PSO) to identify the optimal blend wherein most of the beneficial properties of both the polymers would appear in the resultant blend. In Particle Swarm Optimization, every particle remembers its own previous best value as well as neighborhood best. In this paper, PSO algorithm is used to find the optimal blend ratio (OBR) for cable applications. Also a comparison has been made with Genetic Algorithm (GA) technique, in order to validate the result.