On the Power of Statistical Model Checking

Larsen, Kim Guldstrand; Legay, Axel

doi:10.1007/978-3-319-47169-3_62

Cited by 7 publications

(5 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This work develops an original approach in UPPAAL SMC (David et al, 2015), which enables modelling and analysis of complex adaptive asynchronous multi-agent systems (MAS). Benefits of the approach are formal modelling and the exploitation of statistical model checking techniques (Larsen & Legay, 2016).…”

Section: Discussionmentioning

confidence: 99%

“…Novel in this paper is a mapping of the actor model onto UPPAAL SMC (David et al, 2015) so as to exploit statistical model checking techniques (Younes et al, 2006) (Larsen & Legay, 2016). SMC automatizes multiple executions, estimates a required number of simulation runs, uses statistical properties (e.g., Monte Carlo-like simulations and sequential hypothesis testing) to infer system properties from the observables of the various runs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Statistical Model Checking Of Multi-Agent Systems

Nigro¹,

Sciammarella²

2017

ECMS 2017 Proceedings Edited by Zita Zoltay Paprika, Péter Horák, Kata Váradi, Péter Tamás Zwierczyk, Ágnes Vidovics-Dancs, Ján

View full text Add to dashboard Cite

This paper proposes an original approach to modelling and simulation of multi-agent systems which is based on statistical model checking (SMC). The approach is prototyped in the context of the popular UPPAAL SMC toolbox. Usefulness and validation of the approach are checked by applying it to a known complex and adaptive model of the Iterated Prisoner's Dilemma (IPD) game, by studying the emergence of cooperation in the presence of different social interaction structures.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Statistical Model Checking Of Multi-Agent Systems

Nigro¹,

Sciammarella²

2017

ECMS 2017 Proceedings Edited by Zita Zoltay Paprika, Péter Horák, Kata Váradi, Péter Tamás Zwierczyk, Ágnes Vidovics-Dancs, Ján

View full text Add to dashboard Cite

show abstract

“…Therefore, when the agents are asked to repeat their tasks for multiple rounds, finishing their entire missions becomes a rare event [8], which is hard to achieve when using random simulations. Some algorithms extend statistical model checking, in order to provide a solution to this problem, by using Importance sampling and Importance splitting [19]. Another direction of solving this problem is to integrate the Q-learning algorithm into the process of simulation in UPPAAL.…”

Section: Rare Eventsmentioning

confidence: 99%

Verifiable and Scalable Mission-Plan Synthesis for Autonomous Agents

Enoiu

Seceleanu

et al. 2020

Formal Methods for Industrial Critical Systems

View full text Add to dashboard Cite

The problem of synthesizing mission plans for multiple autonomous agents, including path planning and task scheduling, is often complex. Employing model checking alone to solve the problem might not be feasible, especially when the number of agents grows or requirements include real-time constraints. In this paper, we propose a novel approach called MCRL that integrates model checking and reinforcement learning to overcome this limitation. Our approach employs timed automata and timed computation tree logic to describe the autonomous agents' behavior and requirements, and trains the model by a reinforcement learning algorithm, namely Q-learning, to populate a table used to restrict the state space of the model. Our method provides means to synthesize mission plans for multi-agent systems whose complexity exceeds the scalability boundaries of exhaustive model checking, but also to analyze and verify synthesized mission plans to ensure given requirements. We evaluate the proposed method on various scenarios involving autonomous agents, as well as present comparisons with two similar approaches, TAMAA and UPPAAL STRATEGO. The evaluation shows that MCRL performs better for a number of agents greater than three.

show abstract

“…Using this approach the system will be simulated for a finite number of times.The simulation samples are used for hypothesis testing i.e. to check if the samples provide statistical evidence that the set hypothesis is valid [36]. Two types of tests can be conducted: checking if the probability to satisfy a property is within the specified threshold e.g if the property is voltage deviation ∆V < 5%, and estimation of this probability.…”

Section: Statistical Model Checkingmentioning

confidence: 99%

Analytical Design and Performance Validation of Finite Set MPC Regulated Power Converters

Novak

Nyman

Dragičević

et al. 2019

IEEE Trans. Ind. Electron.

View full text Add to dashboard Cite

A new approach to performance validation of finite control set model predictive control (FCS-MPC) regulated power electronics converters is presented in the paper-statistical model checking (SMC). SMC is an established method used in various sectors of industry to provide solutions to problems that are beyond the abilities of classical formal techniques. The method is simple for implementation and requires only an operational system model that can be simulated and checked against properties. The approach will be presented on a standard 2-level voltage source converter (VSC) regulated by the FCS-MPC algorithm. In UPPAAL SMC toolbox the converter system and controller are modeled as a Network of Stochastic Timed Automata. To assess the quality of the model, an equivalent Simulink model is used as a benchmark model. Using the model created in UPPAAL SMC toolbox the performance of the FCS-MPC algorithm is verified. The control algorithm is also tested on an experimental setup. During the evaluation, no significant degradation of reference tracking was found during transients nor under model parameter uncertainty. Index Terms-Controller performance, DC-AC power conversion, finite control set model predictive control, statistical model checking.

show abstract

On the Power of Statistical Model Checking

Abstract: International audienceThis paper contains material for our tutorial presented at STRESS 2016. This includes an introduction to Statistical Model Checking algorithms and their rare event extensions, as well as an introduction to two well-known SMC tools: Plasma and Uppaal

Cited by 7 publications

References 45 publications

Statistical Model Checking Of Multi-Agent Systems

Statistical Model Checking Of Multi-Agent Systems

Verifiable and Scalable Mission-Plan Synthesis for Autonomous Agents

Analytical Design and Performance Validation of Finite Set MPC Regulated Power Converters

Contact Info

Product

Resources

About