Machinery health prognostics: A systematic review from data acquisition to RUL prediction

Remaining useful life (RUL) prediction allows for predictive maintenance of machinery, thus reducing costly unscheduled maintenance. Therefore, RUL prediction of machinery appears to be a hot issue attracting more and more attention as well as being of great challenge. This paper proposes a model-based method for predicting RUL of machinery. The method includes two modules, i.e., indicator construction and RUL prediction. In the first module, a new health indicator named weighted minimum quantization error is constructed, which fuses mutual information from multiple features and properly correlates to the degradation processes of machinery. In the second module, model parameters are initialized using the maximum-likelihood estimation algorithm and RUL is predicted using a particle filtering-based algorithm. The proposed method is demonstrated using vibration signals from accelerated degradation tests of rolling element bearings. The prediction result identifies the effectiveness of the proposed method in predicting RUL of machinery.

A Model-Based Method for Remaining Useful Life Prediction of Machinery

Remaining useful life estimation using a bidirectional recurrent neural network based autoencoder scheme

Machinery health indicator construction based on convolutional neural networks considering trend burr

Estimation of tire-road friction coefficient based on combined APF-IEKF and iteration algorithm

Health monitoring is an essential part of condition-based maintenance and prognostics and health management for gas turbines. Various health monitoring systems have been developed based on the measurement and observation of the fault symptoms including turbine performance parameters such as heat rate, and nonperformance symptoms such as structural vibration. This paper focuses on surveying state-of-the-art condition monitoring, diagnostic and prognostic techniques using performance parameters acquired from gas-path data that are mostly available from the operating systems of gas turbines. Performance parameters and the corresponding effective factors are presented in the beginning. Structure of performance monitoring and diagnostic systems are systematically laid out next, and the recent developments in each section are surveyed and discussed. Observing the importance of the prognostics in the recent trend of health monitoring research, an emphasis is given on the prognostic frameworks and their implementation for the remaining useful life prediction. A conclusion along with a brief discussion on the current state and potential future directions is provided at the end.

Performance-Based Gas Turbine Health Monitoring, Diagnostics, and Prognostics: A Survey

Performance deterioration monitoring is an essential part of the prognostics and health management (PHM) of gas turbine engines (GTEs). This paper proposes a physics-based modeling approach for performance deterioration monitoring with two model-based performance indicators, heat loss index and power deficit index, for GTE PHM applications. A comprehensive nonlinear thermodynamic model for a single shaft GTE is developed to establish the relation between the operating conditions and the cycle parameters. The model, once properly calibrated, is able to predict the GTE cycle parameters in a healthy condition as the baseline, while in reality, the measured parameters gradually deviate from the baseline, which reflects the performance deterioration of the GTE. To represent the degradation level, the heat loss index is defined as the normalized measure of the thermal power that is being wasted in the GTE compared to the healthy condition. Similarly, the power deficit index is defined as the deficiency ratio of the GTE output power due to the performance deterioration. The effectiveness of the performance indicators in monitoring performance deterioration and their robustness to the variations of the operating conditions are examined by using three years of typical operating data of an industrial GTE. The results clearly reveal the trends of both the short term recoverable deterioration due to fouling effects in the compressor, and the long term non-recoverable deterioration caused by structural degradation. The technique is especially advantageous for prognostic applications where there is no access to internal cycle parameters of a GTE, and only the operating data are available, hence no additional sensors are required.

A Physics-Based Modeling Approach for Performance Monitoring in Gas Turbine Engines

Sequential state estimation of nonlinear/non-Gaussian systems with stochastic input for turbine degradation estimation

Modern health management approaches for gas turbine engines (GTEs) aim to precisely estimate the health state of the GTE components to optimize maintenance decisions with respect to both economy and safety In this research, we propose an advanced framework to identify the most likely degradation state of the turbine section in a GTE for prognostics and health management (PHM) applications A novel nonlinear thermodynamic model is used to predict the performance parameters of the GTE given the measurements The ratio between real efficiency of the GTE and simulated efficiency in the newly installed condition is defined as the health indicator and provided at each sequence The symptom of nonrecoverable degradations in the turbine section, ie loss of turbine efficiency, is assumed to be the internal degradation state A regularized auxiliary particle filter (RAPF) is developed to sequentially estimate the internal degradation state in nonuniform time sequences upon receiving sets of new measurements The effectiveness of the technique is examined using the operating data over an entire time-between-overhaul cycle of a simple-cycle industrial GTE The results clearly show the trend of degradation in the turbine section and the occasional fluctuations, which are well supported by the service history of the GTE The research also suggests the efficacy of the proposed technique to monitor the health state of the turbine section of a GTE by implementing model-based PHM without the need for additional instrumentation

https://iopscience.iop.org/article/10.1088/0957-0233/26/6/065604/pdf

A framework with nonlinear system model and nonparametric noise for gas turbine degradation state estimation

Deposition and congestion of foulants in the compressor section of gas turbine engines (GTE) degrades the compressor and leads to performance deterioration of the GTE at the system level. Compressor fouling may develop over a short time, but it is recoverable by washing and cleaning. Reliable prediction of the fouling as a function of time is helpful for planning compressor wash services. In this work, the fouling state is parametrized as the relative change of the ratio of the compressor mass flow and efficiency against ideal conditions. A regression-based prognostic model is developed to predict the fouling state as a function of time. In the next step, an adaptive neuro-fuzzy inference system (ANFIS) is developed that considers the rate of humidity condensation at the inlet of the compressor for the prognostic model. The performance of the developed models is evaluated with recorded operating data from a GTE in a power plant. The study shows that enhancement of the prognostic model may be accomplished by taking into account the effects of humidity on the rate of fouling and results in an improvement in the prognostic accuracy.

Avisekh Banerjee

Papers

Performance-Based Gas Turbine Health Monitoring, Diagnostics, and Prognostics: A Survey

A Physics-Based Modeling Approach for Performance Monitoring in Gas Turbine Engines

Sequential state estimation of nonlinear/non-Gaussian systems with stochastic input for turbine degradation estimation

A framework with nonlinear system model and nonparametric noise for gas turbine degradation state estimation

Enhancement of prognostic models for short-term degradation of gas turbines