面向图形化编程的事件触发并发数据流模型

Event-triggered concurrent dataflow model for graphic programming

目前的许多图形化编程语言 (又称为可视化语言 )多是基于动态纯数据流 DHDF(dynam ic hom ogeneousdataflow) .由于 DHDF的自然属性 (数据驱动 )与操作系统事件驱动模型不能很好地结合 ,导致了两个明显的不足 :运行效率低 ,CPU占用率高 ;对外部事件响应速度慢 ,系统实时性差 .文中提出了一种事件触发并发数据流 ECDF(event triggered concurrent dataflow)模型 ,并给出了基于该模型的图形化编程语言的构建技术 .ECDF模型通过引入多优先级线程以及事件触发机制 ,在很大程度上提高了系统的实时性与运行效率 ,并且使得基于该模型的图形化编程语言能够实现事件驱动、阻塞调用及重叠 I/O等各种操作模式 .通过一个面向测试系统的图形化程序实例表明 :ECDF模型确实大大提高了图形化程序的运行效率和对外部事件的响应速度 .

Currently, most Graphic Programming Languages (GPL) are based on Dynamic Homogeneous Dataflow (DHDF). Because of its natural data driven property, DHDF cannot work well with the event driven operating system. It leads to inefficienty use of CPU, and low response speed with resulting poor real time performance. An Event Triggered Concurrent Dataflow (ECDF) model and composing technologies of a Graphic Programming Language are presented in this paper. Based on the multi thread and event triggered mechanism, the ECDF can improve the real time performance and execution efficiency of dataflow based systems and enable the GPL to use many driver models such as event driven, blocking call, overlapped I/O, etc. A system test showed that the ECDF greatly enhanced the execution efficiency and response speed of graphic programs.