Minimal Cardinality Diagnosis in Problems with Multiple Observations

Abstract

Model-Based Diagnosis (MBD) is a well-known approach to diagnosis in medical domains. In this approach, the behavior of a system is modeled and used to identify faulty components, i.e., once a symptom of abnormal behavior is observed, an inference algorithm is run on the system model and returns possible explanations. Such explanations are referred to as diagnoses. A diagnosis is an assumption about which set of components are faulty and have caused the abnormal behavior. In this work, we focus on the case where multiple observations are available to the diagnoser, collected at different times, such that some of these observations exhibit symptoms of abnormal behavior. MBD with multiple observations is challenging because some components may fail intermittently, i.e., behave abnormally in one observation and behave normally in another, while other components may fail all the time (non-intermittently). Inspired by recent success in solving classical diagnosis problems using Boolean satisfiability (SAT) solvers, we describe two SAT-based approaches to solve this MBD with multiple observations problem. The first approach compiles the problem to a single SAT formula, and the second approach solves each observation independently and then merges them together. We compare these two approaches experimentally on a standard diagnosis benchmark and analyze their pros and cons.

Model-Based Diagnosis of Multi-Agent Systems: A Survey

“…Furthermore, in some systems, agents may perform faulty actions, but the system does not fail. These are known as Intermittent Faults (Kalech, Stern, and Lazebnik 2021). Observing the MAS in a single execution for a short period of time with such intermittent faults may result in an incorrect diagnosis.…”

Two Compacted Models for Efficient Model-Based Diagnosis

“…The other method to obtain a system description is to generate one formula that encodes the system's knowledge from all the observations (Kalech, Stern, and Lazebnik 2021). This approach finds a solution that enables all the observations to be consistent with the system description by a MaxSAT solver.…”

“…To evaluate our algorithm for an aggregated diagnosis, we compare CMMO and its improvement, D-CMMO, with the state-of-the-art algorithms, namely HSD (Ignatiev et al 2019), IHSD (Zhou et al 2021), DC (Lamraoui and Nakajima 2016), DC* (Ignatiev et al 2019) and One-SAT (Kalech, Stern, and Lazebnik 2021). We use RC2 (Ignatiev et al 2019), a MaxSAT solver, which is also used in the HSD algorithm.…”

“…To the best of our knowledge, there is no standard data sets for MBD problem with multiple observations. In this paper, test cases are generated by mimicking a faulty system as noted in (Ignatiev et al 2019;Kalech, Stern, and Lazebnik 2021). For each circuit, we evaluate the experiment by establishing three fault models: the stuck-at-zero (S-A-0), the stuck-at-one (S-A-1), and flipped fault.…”

“…Its improvement, namely, the Improved implicit Hitting Set Dualization (IHSD) algorithm (Zhou et al 2021), first uses the notion of gate dominance when deducing a diagnosis with multiple observations. The One-SAT algorithm (Kalech, Stern, and Lazebnik 2021) compiles all the observations into a Satisfiability (SAT) formula (Gu et al 1999) and derives the health assignment of all components using a SAT solver. This approach will generate a very large number of clauses when given a large number of observations.…”