摘要
根据物元理论构建超限超重货物装载实例模型,将货物物元分成实例货物物元、实例过程物元、实例方案物元3个子物元,优化装载知识库的结构,在明确实例货物物元相似度计算的基础上,给出基于实例的推理过程,并构造基于实例的装载推理算法,确定新实例的装载参数和装载方案,从而解决了知识获取的瓶颈问题。通过基于机器归纳的自我学习规则算法,补充和扩展实例过程物元的相关推理规则,丰富实例库的实例,进一步提高装载决策和推理过程的智能化水平。
Based on matter element theory, the load instance model of out-of-gauge and super heavy goods has been established to optimize the structure of load knowledge base. In the model, goods matter element is divided into the matter element of instance goods, the matter element of instance process and the matter element of instance scheme. On the basis of ensuring how to calculate the similarity degree of instance goods matter element, the instance-based deduction process is advanced and the load deduction algorithm of out-of-gauge and super heavy goods is constructed in order to determine the load parameters and the load scheme of new instance, and thus the bottleneck problem of knowledge acquisition has been solved. The self-study rule algorithm based on machine deduction is put forward to supplement and extend the relative deduction rules on the matter element of instance process, enrich the instances for instance base and further improve the intelligence level of load decision-making and the deduction process.
出处
《中国铁道科学》
EI
CAS
CSCD
北大核心
2008年第4期116-120,共5页
China Railway Science
基金
铁道部科技研究开发计划项目(2007X012-C)
关键词
超限超重货物
铁路运输
装载决策
实例推理
机器学习
Out-of-gauge and super heavy goods
Railway transportation
Loading decision-making
Instance deduction
Machine learning
