神经计算及其在组合优化中的应用  

Neurocomputation and Its Applications in Combinatorial Optimization

作　　者：陈国良[1] 

机构地区：[1]中国科学技术大学计算机科学技术系,合肥230026

出　　处：《计算机研究与发展》1992年第5期1-21,共21页Journal of Computer Research and Development

摘　　要：本文可分为两大部分:神经计算理论及其在组合优化问题中的应用.前者涉及到计算、数字计算、模拟计算、神经计算的基本概念,计算能量和优化方法;后者列举了旅行商、图的划分、覆盖、最大独立集、最大集团、最大匹配、同构、着色、任务分配、八皇后、模拟数字转换、分解/判定、线性规划和作业调度等组合优化问题的神经网络求解方法.This paper contains two parts:neurocomputing theory and case studies. The former deuls with some basic concepts of the computation, digital computation, analogy computation, biology computation, neurocomputation, computational energy and optimization theory. The latter illustrates some combinatorial optimization problems mapped on to neural networks including traveling salesman problem, graph partitioning problem, graph vertex cover problem, graph independent set problem,graph maximum clique problem, graph maximum matching problem, graph isomorphism problem, task assignment problem, eight queen problem, A/D converter problem, decomposition/decision problem,linear programming problem and job-shop scheduling problem etc.

关 键 词：神经计算 组合优化 神经网络 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]