摘要
在地震走时层析静校正中，反演算法优化一直是个技术难点。与传统的线性反演算法不同，启发式群集智能算法具有自适应、自学习、智能搜索等全局寻优特点，成为一个高效的全局非线性寻优算法。引入量子行为的粒子群优化算法基于概率选择机制，能够有效地克服早熟现象，改善全局搜索能力。在此基础上，将免疫进化算法中的疫苗接种、克隆选择机制引入地震层析成像反演中，以增加抗体的多样性，进一步指导粒子的全局搜索行为，形成了免疫量子粒子群算法。通过理论模型与复杂近地表的静校正资料试算，验证了算法的可行性。
In the course of tomographic static correction of seismic travel-time, the optimization of inversion algorithm has always been technical difficulty. The difference from the conventional linear inversion algorithm is that heuristic colony-forming intellectualized algorithm possesses the following advantages： adaptive, self-studying, intellectualized searching and so on, which has become a high-efficiency overall nonlinear optimization-searching algorithm. The introduction of quantum-behaved particle swarm optimization based on the probability choice mechanism, which can effectively overcome the precocity and improve the overall searching ability. And based on this,the immunization and clone choice mechanism of immunity evolution algorithm are induced into the mapping inversion of seismic tomography to increase antibody＇s diversification and further guide the overall searching of the particles. In the end, the quantum-behaved particle swarm optimization with immunity algorithm （IQPSO） is established. Through the test calculation between the theoretical model and the comprehensive static correction data near the earth surface, the effectiveness and practicability of the algorithm are proved.
出处
《大庆石油地质与开发》
CAS
CSCD
北大核心
2010年第2期130-133,共4页
Petroleum Geology & Oilfield Development in Daqing
基金
中国核工业地质局地勘资金项目（200661）资助.
关键词
量子粒子群优化
免疫量子粒子群优化
免疫接种
克隆选择
Quantum-behaved particle swarm optimization
quantum-behaved particle swarm optimization with immunity algorithm （IQPSO）
immunization
clone choice
