为了解决离散事件系统规范DEVS(discrete event system specification)中没有对智能行为提供直接描述方法的问题,将智能体(Agent)建模思想引入到DEVS中,提出了一种具备一定智能性和协作性描述能力的离散事件系统规范Agent-DEVS。该规范...为了解决离散事件系统规范DEVS(discrete event system specification)中没有对智能行为提供直接描述方法的问题,将智能体(Agent)建模思想引入到DEVS中,提出了一种具备一定智能性和协作性描述能力的离散事件系统规范Agent-DEVS。该规范以并行DEVS为基础,状态元组被扩充成为Agent特征元组;增加了体现个体智能性的Agent模型元组;并将模型端口的输入、输出扩展成为体现社会协作性的Agent消息类型。另外,还给出了Agent-DEVS模型的实现算法,并将其应用于企业生产调度的仿真建模。算法分析与仿真试验表明,Agent-DEVS的优点主要体现在建模性能上,它可以描述更加复杂的智能行为和自治行为,在计算时间上对整体复杂性的影响不大。展开更多
Modelica语言采用微分方程描述系统,此外它还具备面向对象编程语言的特性,因此它不仅适用于连续系统的建模,还支持离散系统的模型架构。因此,可以将Modelica作为一种混合系统的建模语言。提出了一个Modelica语言描述的DEVS(Discrete EVe...Modelica语言采用微分方程描述系统,此外它还具备面向对象编程语言的特性,因此它不仅适用于连续系统的建模,还支持离散系统的模型架构。因此,可以将Modelica作为一种混合系统的建模语言。提出了一个Modelica语言描述的DEVS(Discrete EVent System specification离散事件系统规范)模型架构,并通过对模型的编译过程产生C++代码,获取了同时描述连续系统和离散系统建模的能力。最后给出了用Modelica语言描述的一个飞机导航控制连续-离散仿真系统的例子。展开更多
智能体由于自身交互行为与学习行为的复杂性,难以直接被建模和仿真。针对智能体离散仿真中的常见问题,借助DEVS(discrete event system specification)原子模型的事件转移机制表示智能体的交互与学习过程,通过对智能体交互模式、多状态...智能体由于自身交互行为与学习行为的复杂性,难以直接被建模和仿真。针对智能体离散仿真中的常见问题,借助DEVS(discrete event system specification)原子模型的事件转移机制表示智能体的交互与学习过程,通过对智能体交互模式、多状态外部事件转移控制、端口连接模式、以及强化学习事件转移表示等原子模型下智能体建模技术的介绍,给出了基于DEVS原子模型的智能体离散仿真构建方法。在网格世界与倒立摆2个环境中进行仿真验证,实验结果证明了提出方法在构建智能体交互行为和学习行为的可行性和有效性。展开更多
In Discrete Event System Specification(DEVS),the dynamics of a network is constituted only by the dynamics of its basic components.The state of each component is fully encapsulated.Control in the network is fully dece...In Discrete Event System Specification(DEVS),the dynamics of a network is constituted only by the dynamics of its basic components.The state of each component is fully encapsulated.Control in the network is fully decentralized to each component.At dynamic structure level,DEVS should permit the same level of decentralization.However,it is hard to ensure structure consistency while letting all components achieve structure changes.Besides,this solution can be complex to implement.To avoid these difficulties,usual dynamic structure approaches ensure structure consistency allowing structure changes to be done only by the network having newly added dynamics change capabilities.This is a safe and simple way to achieve dynamic structure.However,it should be possible to simply allow components of a network to modify the structure of their network,other components and/or their own structure—without having to modify the usual definition a DEVS network.In this manuscript,it is shown that a simple fully decentralized approach is possible while ensuring full modularity and structure consistency.展开更多
文摘为了解决离散事件系统规范DEVS(discrete event system specification)中没有对智能行为提供直接描述方法的问题,将智能体(Agent)建模思想引入到DEVS中,提出了一种具备一定智能性和协作性描述能力的离散事件系统规范Agent-DEVS。该规范以并行DEVS为基础,状态元组被扩充成为Agent特征元组;增加了体现个体智能性的Agent模型元组;并将模型端口的输入、输出扩展成为体现社会协作性的Agent消息类型。另外,还给出了Agent-DEVS模型的实现算法,并将其应用于企业生产调度的仿真建模。算法分析与仿真试验表明,Agent-DEVS的优点主要体现在建模性能上,它可以描述更加复杂的智能行为和自治行为,在计算时间上对整体复杂性的影响不大。
文摘Modelica语言采用微分方程描述系统,此外它还具备面向对象编程语言的特性,因此它不仅适用于连续系统的建模,还支持离散系统的模型架构。因此,可以将Modelica作为一种混合系统的建模语言。提出了一个Modelica语言描述的DEVS(Discrete EVent System specification离散事件系统规范)模型架构,并通过对模型的编译过程产生C++代码,获取了同时描述连续系统和离散系统建模的能力。最后给出了用Modelica语言描述的一个飞机导航控制连续-离散仿真系统的例子。
文摘智能体由于自身交互行为与学习行为的复杂性,难以直接被建模和仿真。针对智能体离散仿真中的常见问题,借助DEVS(discrete event system specification)原子模型的事件转移机制表示智能体的交互与学习过程,通过对智能体交互模式、多状态外部事件转移控制、端口连接模式、以及强化学习事件转移表示等原子模型下智能体建模技术的介绍,给出了基于DEVS原子模型的智能体离散仿真构建方法。在网格世界与倒立摆2个环境中进行仿真验证,实验结果证明了提出方法在构建智能体交互行为和学习行为的可行性和有效性。
文摘In Discrete Event System Specification(DEVS),the dynamics of a network is constituted only by the dynamics of its basic components.The state of each component is fully encapsulated.Control in the network is fully decentralized to each component.At dynamic structure level,DEVS should permit the same level of decentralization.However,it is hard to ensure structure consistency while letting all components achieve structure changes.Besides,this solution can be complex to implement.To avoid these difficulties,usual dynamic structure approaches ensure structure consistency allowing structure changes to be done only by the network having newly added dynamics change capabilities.This is a safe and simple way to achieve dynamic structure.However,it should be possible to simply allow components of a network to modify the structure of their network,other components and/or their own structure—without having to modify the usual definition a DEVS network.In this manuscript,it is shown that a simple fully decentralized approach is possible while ensuring full modularity and structure consistency.
