Improve testability of avionics based on optimized Particle Swarm Optimization

This paper provides an optimized PSO (Particle Swarm Optimization) in testability design for complex avionics. It firstly proposes a nonlinear time-varying adjustment strategy which can adjust change rates of learning factors c1 and c2 through variables a and b. Then it simulates different pairs of a and b, aiming at finding out much better change rates of learning factors to optimize the PSO. After that, the optimized PSO will be applied in testability design of complex avionics. Modeling and simulation are taken to see whether the optimized PSO can improve FDR (fault detection rate) and FIR (fault isolation rate). Modeling and simulation suggest that FDR and FIR of the avionics are both improved by the optimized PSO.