1,721,040 research outputs found
Alarm flood reduction using multiple data source
No full textConsulta en la Biblioteca ETSI Industriales (Riunet)[EN] The introduction of distributed control systems in the process industry has increased
the number of alarms per operator exponentially. Modern plants present a high level
of interconnectivity due to steam recirculation, heat integration and the complex
control systems installed in the plant. When there is a disturbance in the plant it
spreads through its material, energy and information connections affecting the process
variables on the path. The alarms associated to these process variables are
triggered. The alarm messages may overload the operator in the control room, who
will not be able to properly investigate each one of these alarms. This undesired situation
is called an “alarm flood”. In such situations the operator might not be able to
keep the plant within safe operation. The aim of this thesis is to reduce alarm flood
periods in process plants. Consequential alarms coming from the same process abnormality
are isolated and a causal alarm suggestion is given. The causal alarm
in an alarm flood is the alarm associated to the asset originating the disturbance
that caused the flood. Multiple information sources are used: an alarm log containing
all past alarms messages, process data and a topology model of the plant. The
alarm flood reduction is achieved with a combination of alarm log analysis, process
data root-cause analysis and connectivity analysis. The research findings are
implemented in a software tool that guides the user through the different steps of
the method. Finally the applicability of the method is proved with an industrial case
study.Rodrigo Marco, JV. (2015). Alarm flood reduction using multiple data source. https://riunet.upv.es/handle/10251/55857.Archivo delegad
Industrial fault detection and diagnosis using alarms and process measurements
Full text linkImplementing data-driven fault detection and diagnosis methods on process plants can be a challenge. Constraints due to the availability and the variability of the process measurements as well as constraints due to the characteristics of the industrial systems impact the reliability of fault detection and diagnosis methods. This thesis aims at increasing the reliability of data-driven fault detection and diagnosis methods on process plants to extend their use in industry.
The core idea of the thesis is to bring together the disciplines of alarm management and fault detection and diagnosis. The first part of the thesis suggests a fault detection and diagnosis approach to the problem of classification of ongoing abnormal situations based on alarm data alone. The second part of the thesis investigates the integration of alarms, alarm settings, and alarm management practices into traditional fault detection and diagnosis methods based on process measurements. Both parts emphasize the robustness of the proposed methods with regard to the variability in the input data, as well as the industrial applicability of the methods. The results are validated on an oil and gas separation plant and on a multiphase flow facility.
The last part of the thesis focuses on root cause analysis of process disturbances. Many data-driven root cause analysis methods have been proposed in process literature in the past twenty years, but their reliability depends on the properties of the industrial system and on the properties of the disturbance. This thesis provides a comparative review of data-driven root cause analysis methods clarifying the scope of application of each method. The objective is to guide practitioners during the root cause analysis and facilitate the use of data-driven root cause analysis methods in industry. The comparative review also highlights the gap of knowledge in root cause analysis of transient disturbances and suggests a new approach based on transient disturbance detection methods to fill this gap.Open Acces
Synthèse d'algorithmes d'optimisation en-ligne par commande extrémale
Full text linkOptimisation en temps réel a base de modèles de connaissance -- Optimisation en temps réel a base de modèles empiriques -- Dépendance de l'erreur sur la solution optimale des méthodes de perturbations a la fréquence d'excitation -- Accélération de la convergence -- Amélioration des propriétés de convergence de la commande extrémale par la méthode des perturbations pour une classe de systèmes différentiellement plats -- Adéquation de modèles pour une optimisation précise via la commande extrémale -- Application a une colonne pilote de flottation
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
A propagation path-based interpretable neural network model for fault detection and diagnosis in chemical process systems
No full text- …
