论文标题

部分可观测时空混沌系统的无模型预测

Automating and Mechanizing Cutoff-based Verification of Distributed Protocols

论文作者

Bhat, Shreesha G., Nagar, Kartik

论文摘要

储层计算是预测湍流的有力工具,其简单的架构具有处理大型系统的计算效率。然而,其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间,然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用,该系统模拟了湍流的若干特征。我们表明,使用径向基函数作为非线性投影器,即使只有部分观测并且不知道控制方程,也能稳健地捕捉复杂的系统非线性。最后,我们表明,当测量稀疏、不完整且带有噪声,甚至控制方程变得不准确时,我们的网络仍然可以产生相当准确的预测,从而为实际湍流系统的无模型预测铺平了道路。

Distributed protocols are generally parametric and can be executed on a system with any number of nodes, and hence proving their correctness becomes an infinite state verification problem. The most popular approach for verifying distributed protocols is to find an inductive invariant which is strong enough to prove the required safety property. However, finding inductive invariants is known to be notoriously hard, and is especially harder in the context of distributed protocols which are quite complex due to their asynchronous nature. In this work, we investigate an orthogonal cut-off based approach to verifying distributed protocols which sidesteps the problem of finding an inductive invariant, and instead reduces checking correctness to a finite state verification problem. The main idea is to find a finite, fixed protocol instance called the cutoff instance, such that if the cutoff instance is safe, then any protocol instance would also be safe. Previous cutoff based approaches have only been applied to a restricted class of protocols and specifications. We formalize the cutoff approach in the context of a general protocol modeling language (RML), and identify sufficient conditions which can be efficiently encoded in SMT to check whether a given protocol instance is a cutoff instance. Further, we propose a simple static analysis-based algorithm to automatically synthesize a cut-off instance. We have applied our approach successfully on a number of complex distributed protocols, providing the first known cut-off results for many of them.

扫码加入交流群

加入微信交流群

微信交流群二维码

扫码加入学术交流群,获取更多资源