【摘要】 提出了一种改进的基于自适应混沌变异的粒子群优化算法来解决地震参数反演问题。该算法提出自适应飞行策略,根据搜索能力对粒子群进行划分,增强了子群间的协同能力,使算法具有良好的全局寻优能力;两阶段混沌变异策略能够在粒子进化的不同阶段进行自适应性搜索,使算法具有较高的搜索精度。实验结果表明,该算法可有效避免标准PSO算法的早熟收敛,具有寻优能力强、搜索精度高、稳定性好等优点。首次将该算法应用于地震参数反演问题,结果表明该算法提高了反演精度且不受初始模型影响,能够较好地解决地震参数反演问题。更多还原

【Abstract】 An improved adaptive PSO algorithm with chaotic mutation(ACMPSO) was proposed. In this algorithm the proposed adaptive flying strategy can result in better global optimum and cooperative capability of the particles by dividing the particle swarms into sub-swarms. Two-step chaotic mutation can adapt the search strategy in different evolution phases. Simulations show that ACMPSO can avoid premature effectively and has powerful optimizing ability, good stability and higher optimizing precision. The proposed ACMPSO is applied into seismic parameters inversion and the results show that the inversion precision is improved greatly and is not influenced by initial model. ACMPSO provides an effective means for seismic parameters inversion.更多还原