[目的/意义]基于文献计量学的新兴技术预测方法能够避免现有德尔菲法的专家依赖性,但是文献所属技术领域的划分仍需依据领域知识设计复杂的检索式。[方法/过程]提出了一种基于机器学习和时间序列预测的新兴技术预测方法,LDA主题模型与S...[目的/意义]基于文献计量学的新兴技术预测方法能够避免现有德尔菲法的专家依赖性,但是文献所属技术领域的划分仍需依据领域知识设计复杂的检索式。[方法/过程]提出了一种基于机器学习和时间序列预测的新兴技术预测方法,LDA主题模型与SVM(支持向量机)分类模型连用的机器学习方法解决了海量论文摘要数据按照新兴技术类别的多分类,通过ARIMA(差分自回归移动平均模型)模型对时序论文未来数量的预测,分析以技术为驱动力的新兴技术发展趋势。[结果/结论]最后,以机器人技术为例,提取Web of Science(WOS)论文数据库上万条数据,对7个新兴领域发展趋势进行预测,为新兴技术预测工作提供有益的数据和方法支持。展开更多
Railway transportation system is a critical sector where design methods and techniques are defined by international standards in order to reduce possible risks to an acceptable minimum level. CENELEC 50128 strongly re...Railway transportation system is a critical sector where design methods and techniques are defined by international standards in order to reduce possible risks to an acceptable minimum level. CENELEC 50128 strongly recommends the utilization of finite state machines during system modelling stage and formal proof methods during the verifi- cation and testing stages of control algorithms. Due to the high importance of interlocking table at the design state of a sig- nalization system, the modelling and verification of inter- locking tables are examined in this work. For this purpose, abstract state machines are used as a modelling tool. The developed models have been performed in a generalized structure such that the model control can be done automatically for the interlocking systems. In this study, NuSMV is used at the verification state. Also, the consistency of the developed models has been supervised through fault injection. The developed models and software components are applied on a real railway station operated by Metro Istanbul Co.展开更多
文摘[目的/意义]基于文献计量学的新兴技术预测方法能够避免现有德尔菲法的专家依赖性,但是文献所属技术领域的划分仍需依据领域知识设计复杂的检索式。[方法/过程]提出了一种基于机器学习和时间序列预测的新兴技术预测方法,LDA主题模型与SVM(支持向量机)分类模型连用的机器学习方法解决了海量论文摘要数据按照新兴技术类别的多分类,通过ARIMA(差分自回归移动平均模型)模型对时序论文未来数量的预测,分析以技术为驱动力的新兴技术发展趋势。[结果/结论]最后,以机器人技术为例,提取Web of Science(WOS)论文数据库上万条数据,对7个新兴领域发展趋势进行预测,为新兴技术预测工作提供有益的数据和方法支持。
文摘Railway transportation system is a critical sector where design methods and techniques are defined by international standards in order to reduce possible risks to an acceptable minimum level. CENELEC 50128 strongly recommends the utilization of finite state machines during system modelling stage and formal proof methods during the verifi- cation and testing stages of control algorithms. Due to the high importance of interlocking table at the design state of a sig- nalization system, the modelling and verification of inter- locking tables are examined in this work. For this purpose, abstract state machines are used as a modelling tool. The developed models have been performed in a generalized structure such that the model control can be done automatically for the interlocking systems. In this study, NuSMV is used at the verification state. Also, the consistency of the developed models has been supervised through fault injection. The developed models and software components are applied on a real railway station operated by Metro Istanbul Co.
