Abnormality Detection and Failure Prediction Using Explainable Bayesian Deep Learning: Methodology and Case Study of Real-World Gas Turbine Anomalies

: Mistrust, amplified by numerous artificial intelligence (AI) related incidents, has caused the energy and industrial sectors to be amongst the slowest adopter of AI methods. Central to this issue is the black-box problem of AI, which impedes investments and fast becoming a legal hazard for users. Explainable AI (XAI) is a recent paradigm to tackle this challenge. Being the backbone of the industry, the prognostic and health management (PHM) domain has recently been introduced to XAI. However, many deficiencies, particularly lack of explanation assessment methods and uncertainty quantification, plague this young field. In this paper, we elaborate a framework on explainable anomaly detection and failure prognostic employing a Bayesian deep learning model to generate local and global explanations from the PHM tasks. An uncertainty measure of the Bayesian model is utilized as marker for anomalies expanding the prognostic explanation scope to include model’s confidence. Also, the global explanation is used to improve prognostic performance, an aspect neglected from the handful of PHM-XAI publications. The quality of the explanation is finally examined employing local accuracy and consistency properties. The method is tested on real-world gas turbine anomalies and synthetic turbofan data failure prediction. Seven out of eight of the tested anomalies were successfully identified. Additionally, the prognostic outcome showed 19% improvement in statistical terms and achieved the highest prognostic score amongst best published results on the topic.

Tackling Variability of Clay to Provide a Robust Binder

Locally available and with infinite recycling possibilities, the use of earth as building material leads to one of the lowest environmental impacts in the construction sector. Recent advances in the earth materials field have been made based on concrete and ceramics technologies to facilitate its uses in dense areas. It is possible to modify clay particle interactions and the material's whole behavior by adding inorganic dispersants and flocculants into clay paste. Earth becomes easy to cast and unmold into formworks, and by removing cement in its composition, poured earth can reach a low CO2 emission rate. Even if this technology is promising, further work has to be performed, as it cannot be implemented on earth from excavation sites with high variability. Tackling the clay nature variability is now the main issue to push this product on the market with robust properties. This research investigates the robustness of the poured earth binder. In this way, several clays (three montmorillonites, two kaolinites, and binary mixes at different proportions) were investigated. Their compacity (C) was determined following the water demand protocol with Vicat apparatus and compared to their consistency properties (liquidity and plasticity limits), and a correlation between these values is established. Different clay pastes prepared at different solid volume fractions were tested to define the influence of the clay nature on the paste consistency evolution. The results showed that clay nature for paste at high solid volume fraction does not influence constituency's evolution when their respectivecompacity is taking into account. It can be suggested that for a clay binder with a consistency close to C, which might be mandatory for poured earth application, only the swelling capacity might influence the mix design.

Decision space robustness for multi-objective integer linear programming

In this article we introduce robustness measures in the context of multi-objective integer linear programming problems. The proposed measures are in line with the concept of decision robustness, which considers the uncertainty with respect to the implementation of a specific solution. An efficient solution is considered to be decision robust if many solutions in its neighborhood are efficient as well. This rather new area of research differs from robustness concepts dealing with imperfect knowledge of data parameters. Our approach implies a two-phase procedure, where in the first phase the set of all efficient solutions is computed, and in the second phase the neighborhood of each one of the solutions is determined. The indicators we propose are based on the knowledge of these neighborhoods. We discuss consistency properties for the indicators, present some numerical evaluations for specific problem classes and show potential fields of application.

On Probabilistic Logical Argumentation

In this paper, we introduce a framework for probabilistic logic-based argumentation inspired on the DeLP formalism and an extensive use of conditional probability. We define probabilistic arguments built from possibly inconsistent probabilistic knowledge bases and study the notions of attack, defeat and preference between these arguments. Finally, we discuss consistency properties of admissible extensions of the Dung’s abstract argumentation graphs obtained from sets of probabilistic arguments and the attack relations between them.

Consistency properties for the wavelet estimator in nonparametric regression model with dependent errors

In this paper, we establish the pth mean consistency, complete consistency, and the rate of complete consistency for the wavelet estimator in a nonparametric regression model with m-extended negatively dependent random errors. We show that the best rates can be nearly $O(n^{-1/3})$ O ( n − 1 / 3 ) under some general conditions. The results obtained in the paper markedly improve and extend some corresponding ones to a much more general setting.

Scalable Reaction Network Modeling with Automatic Validation of Consistency in Event-B

Constructing a large biological model is a difficult, error-prone process. Small errors in writing a part of the model cascade to the system level and their sources are difficult to trace back. In this paper we extend a recent approach based on Event-B, a state-based formal method with refinement as its central ingredient, allowing us to validate for model consistency step-by-step in an automated way. We demonstrate this approach on a model of the heat shock response and its scalability on a model of the ErbB signaling pathway, a key evolutionary pathway with a significant role in development and in many types of cancer. All consistency properties of the model were proved automatically with computer support.

On sectoriality of degenerate elliptic operators

Let $c_{kl} \in W^{1,\infty }(\Omega , \mathbb{C})$ for all $k,l \in \{1, \ldots , d\};$ and $\Omega \subset \mathbb{R}^{d}$ be open with uniformly $C^{2}$ boundary. We consider the divergence form operator $A_p = - \sum \nolimits _{k,l=1}^{d} \partial _l (c_{kl} \partial _k)$ in $L_p(\Omega )$ when the coefficient matrix satisfies $(C(x) \xi , \xi ) \in \Sigma _\theta$ for all $x \in \Omega$ and $\xi \in \mathbb{C}^{d}$ , where $\Sigma _\theta$ be the sector with vertex 0 and semi-angle $\theta$ in the complex plane. We show that a sectorial estimate holds for $A_p$ for all $p$ in a suitable range. We then apply these estimates to prove that the closure of $-A_p$ generates a holomorphic semigroup under further assumptions on the coefficients. The contractivity and consistency properties of these holomorphic semigroups are also considered.

THE DEVELOPMENT OF BISCHOFITE ANTI-SCAR GEL COMPOSITION

The development of an external gel containing purified bischofite will allow the scars treatment in the stage of prevention and formation due to the effect on various links of pathological wound healing. This study’s aim was the development of a gel with bischofite for the scars prevention and treatment. Bischofite brine from the purified Volgograd deposit was selected as the active pharmaceutical substance. Polymer gelling agents: methylcellulose-100, sodium carboxymethyl cellulose, hydroxymethyl cellulose, aerosil, Tizol®. The QTPP requirements for developed bischofite gel are aimed at effective wound healing and prevention of pathological scarring, which corresponds to the ointments used at the III stage of the wound process. Comprehensive technological studies of model samples of gels with bischofite were carried out: determination of external signs and the application to the skin, smearing, thermal stability and pH, study of osmotic activity and release of bischofite (in terms of magnesium ions). The maximum amount (8 points) was observed in the composition using the Tizol® gel-forming agent, the model sample based on it provided the maximum degree of release, minimal osmotic activity, and good smearing. In addition, Tizol® possesses anti-inflammatory activity. The optimal concentration of the aqueous phase is justified by the assessment of the consistency properties, the spreadability and rheological properties. Thus, as a result the composition of the bischofite gel was developed using Tizol® as a base, containing glycerin as a plasticizer and a moisture agent, the preservative sodium benzoate and purified water.

An Application of an Embedded Model Estimator to a Synthetic Nonstationary Reservoir Model With Multiple Secondary Variables

A method (Ember) for nonstationary spatial modeling with multiple secondary variables by combining Geostatistics with Random Forests is applied to a three-dimensional Reservoir Model. It extends the Random Forest method to an interpolation algorithm retaining similar consistency properties to both Geostatistical algorithms and Random Forests. It allows embedding of simpler interpolation algorithms into the process, combining them through the Random Forest training process. The algorithm estimates a conditional distribution at each target location. The family of such distributions is called the model envelope. An algorithm to produce stochastic simulations from the envelope is demonstrated. This algorithm allows the influence of the secondary variables, as well as the variability of the result to vary by location in the simulation.

UNIFORM ASYMPTOTICS AND CONFIDENCE REGIONS BASED ON THE ADAPTIVE LASSO WITH PARTIALLY CONSISTENT TUNING

We consider the adaptive Lasso estimator with componentwise tuning in the framework of a low-dimensional linear regression model. In our setting, at least one of the components is penalized at the rate of consistent model selection and certain components may not be penalized at all. We perform a detailed study of the consistency properties and the asymptotic distribution which includes the effects of componentwise tuning within a so-called moving-parameter framework. These results enable us to explicitly provide a set $\mathcal {M}$ such that every open superset acts as a confidence set with uniform asymptotic coverage equal to 1, whereas removing an arbitrarily small open set along the boundary yields a confidence set with uniform asymptotic coverage equal to 0. The shape of the set $\mathcal {M}$ depends on the regressor matrix as well as the deviations within the componentwise tuning parameters. Our findings can be viewed as a broad generalization of Pötscher and Schneider (2009, Journal of Statistical Planning and Inference 139, 2775–2790; 2010, Electronic Journal of Statistics 4, 334–360), who considered distributional properties and confidence intervals based on components of the adaptive Lasso estimator for the case of orthogonal regressors.