Implementing Quantum-Behaved Particle Swarm Optimization Algorithm in FPGA for Embedded Real-Time Applications
Zhilei Chai, Jun Jie Sun, Rui Cai, Wenbo Xu · 2009
Quantum-behaved Particle Swarm Optimization (QPSO) algorithm is the algorithm proposed by our group to improve traditional PSOs. Comparing with traditional PSOs, the QPSO has better global-convergence and fewer parameters. Both PSO and QPSO were generally considered to have high requirement for the computing platform; hence, it is thought impractical to employ them into resource-constrained embedded real-time systems. However, with the development of FPGA technology, recently many complex algorithms were implemented in FPGA to get high computing performance for the embedded real-time applications. In this paper, a microarchitecture specific for QPSO suited to be implemented in a low-end FPGA is proposed. Furthermore, this microarchitecture is designed with reconfigurability to meet different requirements on computing performance and resource consumption.