Michael E. Cholette

Monitoring of complex systems of interacting dynamic systems

Increases in functionality, power and intelligence of modern engineered systems led to complex systems with a large number of interconnected dynamic subsystems. In such machines, faults in one subsystem can cascade and affect the behavior of numerous other subsystems. This complicates the traditional fault monitoring procedures because of the need to train models of the faults that the monitoring system needs to detect and recognize. Unavoidable design defects, quality variations and different usage patterns make it infeasible to foresee all possible faults, resulting in limited diagnostic coverage that can only deal with previously anticipated and modeled failures. This leads to missed detections and costly blind swapping of acceptable components because of one’s inability to accurately isolate the source of previously unseen anomalies. To circumvent these difficulties, a new paradigm for diagnostic systems is proposed and discussed in this paper. Its feasibility is demonstrated through application examples in automotive engine diagnostics.

Precedent-Free Fault Isolation in a Diesel Engine Exhaust Gas Recirculation System

In this paper, a recently introduced model-based method for precedent-free fault detection and isolation (FDI) is modified to deal with multiple input, multiple output (MIMO) systems and is applied to an automotive engine with exhaust gas recirculation (EGR) system. Using normal behavior data generated by a high fidelity engine simulation, the growing structure multiple model system (GSMMS) approach is used to construct dynamic models of normal behavior for the EGR system and its constituent subsystems. Using the GSMMS models as a foundation, anomalous behavior is detected whenever statistically significant departures of the most recent modeling residuals away from the modeling residuals displayed during normal behavior are observed. By reconnecting the anomaly detectors (ADs) to the constituent subsystems, EGR valve, cooler, and valve controller faults are isolated without the need for prior training using data corresponding to particular faulty system behaviors.

Precedent-Free Fault Isolation in a Diesel Engine EGR Valve System

In this paper, a framework for isolating unprecedented faults for an EGR valve system is presented. Using normal behavior data generated by a high fidelity engine simulation, the recently introduced Growing Structure Multiple Model System (GSMMS) is used to construct models of normal behavior for an EGR valve system and its various subsystems. Using the GSMMS models as a foundation, anomalous behavior of the entire system is then detected as statistically significant departures of the most recent modeling residuals from the modeling residuals during normal behavior. By reconnecting anomaly detectors to the constituent subsystems, the anomaly can be isolated without the need for prior training using faulty data. Furthermore, faults that were previously encountered (and modeled) are recognized using the same approach as the anomaly detectors.

A Dual-Mode Model Predictive Control Algorithm Trajectory Tracking in Discrete-Time Nonlinear Dynamic Systems

In this paper, a dual-mode model predictive/linear control method is presented, which extends the concept of dual-mode model predictive control (MPC) to trajectory tracking control of nonlinear dynamic systems described by discrete-time state-space models. The dual-mode controller comprises of a time-varying linear control law, implemented when the states lie within a sufficiently small neighborhood of the reference trajectory, and a model predictive control strategy driving the system toward that neighborhood. The boundary of this neighborhood is characterized so as to ensure stability of the closed-loop system and terminate the optimization procedure in a finite number of iterations, without jeopardizing the stability of the closed-loop system. The developed controller is applied to the central air handling unit (AHU) of a two-zone variable air volume (VAV) heating, ventilation, and air conditioning (HVAC) system.

Extracting failure time data from industrial maintenance records using text mining

Reliability modelling requires accurate failure time of an asset. In real industrial cases, such data are often buried in different historical databases which were set up for purposes other than reliability modelling. In particular, two data sets are commonly available: work orders (WOs), which detail maintenance activities on the asset, and downtime data (DD), which details when the asset was taken offline. Each is incomplete from a failure perspective, where one wishes to know whether each downtime event was due to failure or scheduled activities. In this paper, a text mining approach is proposed to extract accurate failure time data from WOs and DD. A keyword dictionary is constructed using WO text descriptions and classifiers are constructed and applied to attribute each of the DD events to one of two classes: failure or nonfailure. The proposed method thus identifies downtime events whose descriptions are consistent with urgent unplanned WOs. The applicability of the methodology is demonstrated on maintenance data sets from an Australian electricity and sugar processing companies. Analysis of the text of the identified failure events seems to confirm the accurate identification of failures in DD. The results are expected to be immediately useful in improving the estimation of failure times (and thus the reliability models) for real-world assets.

Using support vector machines for the computationally efficient identification of acceptable design parameters in computer-aided engineering applications

SVMs are used to estimate the boundary of acceptable design parameters.An active learning method is developed to efficiently refine the boundary estimate.The algorithm is applied to a (known) toy function to demonstrate its effectiveness.The approach is subsequently used to find the dynamic stability limit of a train. This paper addresses the problem of estimating continuous boundaries between acceptable and unacceptable engineering design parameters in complex engineering applications. In particular, a procedure is proposed to reduce the computational cost of finding and representing the boundary. The proposed methodology combines a low-discrepancy sequence (Sobol) and a support vector machine (SVM) in an active learning procedure able to efficiently and accurately estimate the boundary surface. The paper describes the approach and methodological choices resulting in the desired level of boundary surface refinement and the new algorithm is applied to both two highly-nonlinear test functions and a real-world train stability design problem. It is expected that the new method will provide designers with a tool for the evaluation of the acceptability of designs, particularly for engineering systems whose behaviour can only be determined through complex simulations.

An investigation on factors influencing dust accumulation on CSP mirrors

The profitability of a CSP plant is highly affected by the efficiency of the solar field: it is essential to maintain mirrors’ reflectivity at high level to avoid thermal power loss. Dust fouling is the main cause of reflectivity loss and cleaning of mirrors is a crucial activity to restore economical level of reflectivity. However, the high cost of cleaning operations requires the study and identification of a balanced plan for the dust removal. The dust generation and transport to the plant site is the first mechanism that needs to be modelled to identify the optimal schedule for cleaning operations and it is highly dependent on weather conditions. Several studies have suggested a dependency of reflectors performance with humidity level, frequency of rainfalls, wind and mirrors’ tilting angle, however rarely quantitative correlation studies have been performed to validate these hypotheses. The aim of this research is to provide an in-depth insight on interaction between the main parameters and airborne ...

Predicting Maintenance Requirements for School Assets in Queensland

In this paper, a maintenance prediction model is developed for school building assets using a large data set provided by the Queensland Department of Education and Training (DET). DET data on the asset condition, historical maintenance expenditure, and asset characteristics, was analyzed to evaluate which characteristics affect the maintenance needs of the school assets. The condition of the assets was quantified using data on the estimated maintenance backlog. Using statistical methods, models for key building element groups were constructed and the statistical significance of each factor was evaluated. It was found that the school region, the gross floor area, and the maintenance expenditure significantly affected the degradation of key building element groups.

Energy Efficiency and Tracking Performance Evaluation for Dual-Mode Model Predictive Control of HVAC Systems

This paper presents evaluation of the energy consumption and tracking performance associated with the use of a recently introduced dual-mode model predictive controller (DMMPC) for control of a heating, ventilation, and air conditioning (HVAC) system. The study was conducted using detailed simulations of an HVAC system, which included a multizone air loop, a water loop, and a chiller. Energy consumption and tracking performance are computed from the simulations and evaluated in the presence of different types and magnitudes of noise and disturbances. Performance of the DMMPC is compared with a baseline proportional-integral-derivative (PID) control structure commonly used for HVAC system control, and this comparison showed clear and consistent superiority of the DMMPC.

Opportunistic Maintenance for Wind Turbines Considering External Opportunities – A Case Study

This paper aims to develop an opportunistic maintenance (OM) policy for the generator of a hypothetical wind turbine using methods developed recently by the authors. The OM policy considers external opportunities caused by low wind speeds which produce little-to-no electric power. The results show that some cost savings are achievable by taking maximal advantage of these low-speed wind events, particularly when electricity prices are at their peak cycle.