Particle swarm optimization （PSO） algorithm is an effective bio-inspired algorithm but it has shortage of premature convergence. Researchers have made some improvements especially in force rules and population topol...Particle swarm optimization （PSO） algorithm is an effective bio-inspired algorithm but it has shortage of premature convergence. Researchers have made some improvements especially in force rules and population topologies. However, the current algorithms only consider a single kind of force rules and lack consideration of comprehensive improvement in both multi force rules and population topologies. In this paper, a dynamic topology multi force particle swarm optimization （DTMFPSO） algorithm is proposed in order to get better search performance. First of all, the principle of the presented multi force particle swarm optimization （MFPSO） algorithm is that different force rules are used in different search stages, which can balance the ability of global and local search. Secondly, a fitness-driven edge-changing （FE） topology based on the probability selection mechanism of roulette method is designed to cut and add edges between the particles, and the DTMFPSO algorithm is proposed by combining the FE topology with the MFPSO algorithm through concurrent evolution of both algorithm and structure in order to further improve the search accuracy. Thirdly, Benchmark functions are employed to evaluate the performance of the DTMFPSO algorithm, and test results show that the proposed algorithm is better than the well-known PSO algorithms, such as gPSO, MPSO, and EPSO algorithms. Finally, the proposed algorithm is applied to optimize the process parameters for ultrasonic vibration cutting on SiC wafer, and the surface quality of the SiC wafer is improved by 12.8% compared with the PSO algorithm in Ref. [25]. This research proposes a DTMFPSO algorithm with multi force rules and dynamic population topologies evolved simultaneously, and it has better search performance.展开更多
Existing multifactorial particle swarm optimisation(MFPSO)algorithms only explore a relatively narrow area between the inter-task particles.Meanwhile,these algorithms use a fixed inter-task learning probability throug...Existing multifactorial particle swarm optimisation(MFPSO)algorithms only explore a relatively narrow area between the inter-task particles.Meanwhile,these algorithms use a fixed inter-task learning probability throughout the evolution process.However,the parameter is problem dependent and can be various at different stages of the evolution.In this work,the authors devise an inter-task learning-based information transferring mechanism to replace the corresponding part in MFPSO.This inter-task learning mechanism transfers the searching step by using a differential term and updates the personal best position by employing an inter-task crossover.By this mean,the particles can explore a broad search space when utilising the additional searching experiences of other tasks.In addition,to enhance the performance on problems with different complementarity,they design a self-adaption strategy to adjust the inter-task learning probability according to the performance feedback.They compared the proposed algorithm with the state-of-the-art algorithms on various benchmark problems.Experimental results demonstrate that the proposed algorithm can transfer inter-task knowledge efficiently and perform well on the problems with different complementarity.展开更多
基金Supported by National Natural Science Foundation of China(Grant No.51405426)Hebei Provincial Natural Science Foundation of China(Grant No.E2016203306)
文摘Particle swarm optimization （PSO） algorithm is an effective bio-inspired algorithm but it has shortage of premature convergence. Researchers have made some improvements especially in force rules and population topologies. However, the current algorithms only consider a single kind of force rules and lack consideration of comprehensive improvement in both multi force rules and population topologies. In this paper, a dynamic topology multi force particle swarm optimization （DTMFPSO） algorithm is proposed in order to get better search performance. First of all, the principle of the presented multi force particle swarm optimization （MFPSO） algorithm is that different force rules are used in different search stages, which can balance the ability of global and local search. Secondly, a fitness-driven edge-changing （FE） topology based on the probability selection mechanism of roulette method is designed to cut and add edges between the particles, and the DTMFPSO algorithm is proposed by combining the FE topology with the MFPSO algorithm through concurrent evolution of both algorithm and structure in order to further improve the search accuracy. Thirdly, Benchmark functions are employed to evaluate the performance of the DTMFPSO algorithm, and test results show that the proposed algorithm is better than the well-known PSO algorithms, such as gPSO, MPSO, and EPSO algorithms. Finally, the proposed algorithm is applied to optimize the process parameters for ultrasonic vibration cutting on SiC wafer, and the surface quality of the SiC wafer is improved by 12.8% compared with the PSO algorithm in Ref. [25]. This research proposes a DTMFPSO algorithm with multi force rules and dynamic population topologies evolved simultaneously, and it has better search performance.
文摘Existing multifactorial particle swarm optimisation(MFPSO)algorithms only explore a relatively narrow area between the inter-task particles.Meanwhile,these algorithms use a fixed inter-task learning probability throughout the evolution process.However,the parameter is problem dependent and can be various at different stages of the evolution.In this work,the authors devise an inter-task learning-based information transferring mechanism to replace the corresponding part in MFPSO.This inter-task learning mechanism transfers the searching step by using a differential term and updates the personal best position by employing an inter-task crossover.By this mean,the particles can explore a broad search space when utilising the additional searching experiences of other tasks.In addition,to enhance the performance on problems with different complementarity,they design a self-adaption strategy to adjust the inter-task learning probability according to the performance feedback.They compared the proposed algorithm with the state-of-the-art algorithms on various benchmark problems.Experimental results demonstrate that the proposed algorithm can transfer inter-task knowledge efficiently and perform well on the problems with different complementarity.
