Boundary Element Inverse Analysis by Using Particle Swarm Optimization for Reinforced Concrete Corrosion Identification

Abstract

Boundary element inverse analysis (BEIA) by using genetic algorithm (GA) to identify corrosion location has been introduced by many researchers. However, the BEIA using GA is more complex to be programmed since it involved with genetic operators such as crossover and mutation. Recently, Particle Swarm Optimization (PSO) already takes researcher’s attention because of its simplicity to be programmed and comparable accuracy. This study is conducted to develop BEIA by combining Boundary Element Method (BEM) and PSO to identify the corrosion location of the steels in concrete structure from some potential data on concrete surface. The potential in the concrete domain was modeled by Laplace’s equation. The anode and cathode are represented by each polarization curve. The inverse problem is carried out by means of minimizing a cost function i.e. a difference between the calculated and measured potentials on the concrete surface. The calculated values of potential are obtained by solving the Laplace’s equation using boundary element method (BEM). Numerical simulation results show that the developed BEIA has proven that it can identify the corrosion location on the surface of reinforcement steel precisely.