Search arXivSearch

arXiv · 1301.0039

Generating Property-Directed Potential Invariants By Backward Analysis

Abstract

This paper addresses the issue of lemma generation in a k-induction-based formal analysis of transition systems, in the linear real/integer arithmetic fragment. A backward analysis, powered by quantifier elimination, is used to output preimages of the negation of the proof objective, viewed as unauthorized states, or gray states. Two heuristics are proposed to take advantage of this source of information. First, a thorough exploration of the possible partitionings of the gray state space discovers new relations between state variables, representing potential invariants. Second, an inexact exploration regroups and over-approximates disjoint areas of the gray state space, also to discover new relations between state variables. k-induction is used to isolate the invariants and check if they strengthen the proof objective. These heuristics can be used on the first preimage of the backward exploration, and each time a new one is output, refining the information on the gray states. In our context of critical avionics embedded systems, we show that our approach is able to outperform other academic or commercial tools on examples of interest in our application field. The method is introduced and motivated through two main examples, one of which was provided by Rockwell Collins, in a collaborative formal verification framework.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Adrien Champion, Rémi Delmas, Michael Dierkes. 2013-01-01. Generating Property-Directed Potential Invariants By Backward Analysis. https://doi.org/10.4204/eptcs.105.3

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

A Probabilistic Choreography Language for PRISM

We present a choreographic framework for modelling and analysing concurrent probabilistic systems based on the PRISM model-checker. This is achieved through the development of a choreography language, which is a specification language that allows to describe the desired interactions within a concurrent system from a global viewpoint. Using choreographies gives a clear and complete view of system interactions, making it easier to understand the process flow and identify potential errors, which helps ensure correct execution and improves system reliability. We equip our language with a probabilistic semantics and then define a formal encoding into the PRISM language and discuss its correctness. Properties of programs written in our choreographic language can be model-checked by the PRISM model-checker via their translation into the PRISM language. Finally, we implement a compiler for our language and demonstrate its practical applicability via examples drawn from the use cases featured in the PRISM website.

cs.LO

Monotone weak distributive laws over the lifted powerset monad in categories of algebras

Noticing the similarity between the monotone weak distributive laws combining two layers of nondeterminism in sets and in compact Hausdorff spaces, we study whether the latter law can be obtained automatically as a weak lifting of the former. This holds partially, but does not generalize to other categories of algebras: we then characterize when exactly monotone weak distributive laws over powerset monads in categories of algebras exist, exhibiting a law combining probabilities and non-determinism in compact Hausdorff spaces and showing on the other hand that such laws do not exist in a lot of other cases.

cs.LO

Identifying and Explaining (Non-)Equivalence of First-Order Logic Formulas

First-order logic is the basis for many knowledge representation formalisms and methods. Providing technological support for learning to write first-order formulas for natural language specifications requires methods to test formulas for (non-)equivalence and to provide explanations for non-equivalence. We propose such methods based on both theoretical insights and existing tools, implement them, and report on experiments testing their effectiveness on a large educational data set with > 100,000 pairs of first-order formulas.

cs.LO