Lijie Yu

A novel approach to aircraft engine anomaly detection and diagnostics

Accurate and timely failure detection and diagnosis is critical to reliable and affordable aircraft engine operation. This work describes a statistical and fuzzy logic based approach that analyzes multiple engine performance parameters for trend recognition, shift evaluation and failure classification. It integrates the statistical data analysis and fuzzy logic reasoning processes and provides powerful data fusion capability. The system captures and diagnoses failures as soon as the engine performance-shifting trend is recognizable, based on customizable probability. This approach improves upon current diagnostic processes in a number of ways. First, the dimensionality is increased so that multiple relevant parameters are integrated into the diagnosis. This helps reduce single dimension false alarms. Second, this approach effectively handles the noise in engine performance data. Many diagnoses depend on detecting changes in the data that fall within three standard deviations of the pre-event data, historically leading to false alerts and diagnoses. Finally, this approach seamlessly integrates the noise in the data with the uncertainty in the diagnostic models, rolling it up into a single score for each potential diagnosis. This increases consistency, and removes a substantial amount of subjective judgment from the diagnostic process. This approach has been successfully applied to a series of General Electric commercial airline engines, demonstrating high accuracy and consistency. The methodology is expected to be generally applicable to a wide variety of engine models and failure modes.

Information Fusion Strategy for Aircraft Engine Health Management

Modern aircraft engines are equipped with sophisticated sensing instruments to enable proactive condition monitoring and effective health management capability. Development of intelligent systems that efficiently process sensor and operational data both onboard and off-board, to provide maintenance personnel with timely decision support, is the key to minimize flight service disruption and reduce engine ownership cost. The goal of this research is to develop a practical approach and strategy to leverage various available information sources and modeling techniques to streamline the engine health management process and maximize system accuracy and efficiency. This paper demonstrates a flexible fusion architecture that encapsulates the key elements of the engine monitoring and diagnostic process, i.e., sensor trend analysis module for anomaly detection, feature selection and fault isolation module for root cause identification, a decision module for diagnostic model fusion and action determination, and finally, a feedback module for knowledge validation and continuous learning. At the core of this engine health management system is a diagnostic fusion model designed around a common fault hierarchy which captures both a priori probabilities and interactions among various engine faults isolated by different classification models. The fusion model will resolve conflicting assessments from individual diagnostic models and provide a more accurate and comprehensive engine state estimate.© 2007 ASME

Decision Support for Remote Monitoring and Diagnostics of Aircraft Engine Using Influence Diagrams

FAA regulations require the monitoring of all commercial aircraft engines to ensure airworthiness. In doing so, it provides economic advantages to engine owners to monitor engine performance and resolve identified issues in a timely manner to reduce operational costs or avoid secondary damage. Various remote monitoring and diagnostics service providers exist in the marketplace. However, a common understanding among most of them is that given limited time and information, it is an extremely difficult task to make quick and optimized decisions. Difficulties arise from the fact that an aircraft engine is a complex system and demands considerable expertise to diagnose, but also due to the uncertainty in estimating an engine’s true physical state because of measurement and process noise. Therefore, it is often difficult to decide what action to take in order to achieve the most desirable outcome. In this paper, a cost sensitive engine diagnostic and decision making methodology is described. Diagnostic tool performance at various decision thresholds is estimated over a large set of validated historical cases to evaluate sensitivity, specificity and other quality indices. These quality indices and a set of cost functions are utilized in influence diagrams to derive the optimized decision model in order to minimize costs given the uncertain engine condition and noisy parametric data.Copyright