基于量子粒子群优化BP网络的计算机层析成像反演算法  

CT Inversion Algorithm Based on BP Neural Network Optimized by Quantum Particle Swarm Optimization

作　　者：刘李楠 LIU Li-nan(Naval Petty Officer Academy,Bengbu 233012,Anhui)

机构地区：[1]海军士官学校,安徽蚌埠233012

出　　处：《电脑与电信》2021年第6期57-59,64,共4页Computer & Telecommunication

摘　　要：计算机层析成像技术在很多领域有着广泛应用,其中反演算法的优劣将直接关系层析成像的成败。提出一种基于量子粒子群优化BP网络的计算机层析成像反演算法,利用量子粒子群算法优化BP神经网络的权值和阈值,优化训练后的网络能更好完成大型稀疏非线性方程组求解问题。仿真实验表明方法具有较强的实用性及计算效率。Computer tomography is widely used in many fields,and the quality of inversion algorithm is directly related to the suc-cess of tomography.In this paper,a computer tomography inversion algorithm based on BP neural network optimized by Quantum Particle Swarm Optimization(QPSO)is proposed.The weights and thresholds of BP neural network are optimized by QPSO.The trained network can better solve large sparse nonlinear equations.The simulation result show that the method is practical and effi-cient.

关 键 词：层析成像 反演算法 量子粒子群 BP网络 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]