摘要
采用人工神经网络中的基本Hopfield网,实现了空间机器人的路径规划,解决了对两两给定通路的若干空间站点,空间机器人能顺序访问并最终返回起点的最短巡游路径问题.用Hopfield网解决最优化问题可针对不同的具体问题来设计网络的连接权及能量函数,从而扩展网络对解决具体问题的适应性,具有并行处理能力强、方法简单、容易实现等诸多优点.
A path planning for space robot by using basic Hopfield network is presented, which gives the solution of finding the shortest path for space robot to call on every space station and return. Connection weighting and energy function can be designed according to every actual problem so as to extend the network adaptability for solving actual problems. The Hopfield network has many advantages, such as the massively parallel processing abilities as well as simplicity in method use.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2003年第8期970-973,共4页
Journal of Harbin Institute of Technology
基金
国家高技术研究发展计划资助项目(863-2-4-1-2).