A Genetic Algorithm Optimized RNN-LSTM Model for Remaining Useful Life Prediction of Turbofan Engine

Understanding the remaining useful life (RUL) of equipment is crucial for optimal predictive maintenance (PdM). This addresses the issues of equipment downtime and unnecessary maintenance checks in run-to-failure maintenance and preventive maintenance. Both feature extraction and prediction algorithm have played crucial roles on the performance of RUL prediction models. A benchmark dataset, namely Turbofan Engine Degradation Simulation Dataset, was selected for performance analysis and evaluation. The proposal of the combination of complete ensemble empirical mode decomposition and wavelet packet transform for feature extraction could reduce the average root-mean-square error (RMSE) by 5.14–27.15% compared with six approaches. When it comes to the prediction algorithm, the results of the RUL prediction model could be that the equipment needs to be repaired or replaced within a shorter or a longer period of time. Incorporating this characteristic could enhance the performance of the RUL prediction model. In this paper, we have proposed the RUL prediction algorithm in combination with recurrent neural network (RNN) and long short-term memory (LSTM). The former takes the advantages of short-term prediction whereas the latter manages better in long-term prediction. The weights to combine RNN and LSTM were designed by non-dominated sorting genetic algorithm II (NSGA-II). It achieved average RMSE of 17.2. It improved the RMSE by 6.07–14.72% compared with baseline models, stand-alone RNN, and stand-alone LSTM. Compared with existing works, the RMSE improvement by proposed work is 12.95–39.32%.

Download Full-text

Remaining Useful Life Prediction of Turbofan Engine Using Hybrid Model Based on Autoencoder and Bidirectional Long Short-Term Memory

Journal of Shanghai Jiaotong University (Science) ◽

10.1007/s12204-018-2027-5 ◽

2018 ◽

Vol 23 (S1) ◽

pp. 85-94 ◽

Cited By ~ 4

Author(s):

Ya Song ◽

Guo Shi ◽

Leyi Chen ◽

Xinpei Huang ◽

Tangbin Xia

Keyword(s):

Hybrid Model ◽

Life Prediction ◽

Short Term Memory ◽

Remaining Useful Life ◽

Short Term ◽

Turbofan Engine ◽

Term Memory ◽

Model Based ◽

Useful Life ◽

Long Short Term Memory

Download Full-text

Remaining Useful Life Prediction of Turbofan Engine using Long-Short Term Memory

10.1109/3ict53449.2021.9581576 ◽

2021 ◽

Author(s):

Faatih Nuraliah Binti Sohaidan ◽

Amgad Muneer ◽

Shakirah Mohd Taib

Keyword(s):

Life Prediction ◽

Short Term Memory ◽

Remaining Useful Life ◽

Short Term ◽

Turbofan Engine ◽

Term Memory ◽

Useful Life ◽

Long Short Term Memory

Download Full-text

Short-Term Prediction Model for a Grid-Connected Photovoltaic System Using EMD and GABPNN

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.291-294.74 ◽

2013 ◽

Vol 291-294 ◽

pp. 74-82

Author(s):

Zeng Wei Zheng ◽

Yuan Yi Chen ◽

Xiao Wei Zhou ◽

Mei Mei Huo ◽

Bo Zhao ◽

...

Keyword(s):

Neural Network ◽

Genetic Algorithm ◽

Prediction Model ◽

Prediction Method ◽

Photovoltaic System ◽

Photovoltaic Systems ◽

Short Term ◽

Model Based ◽

Term Prediction ◽

Short Term Prediction

The integration between photovoltaic systems and tradition grid have a lot of challenges. To accurately predict is a key to solve these challenges. Due to complex, non-linear and non-stationary characteristics, it is difficult to accurately predict the power of photovoltaic systems. In this paper, a short-term prediction model based on empirical mode decomposition (EMD)and back propagation neural network(BPNN) was constructed, and use genetic algorithm as the learn algorithm of BPNN. The power data after pre-processing is decomposed into several components, then using prediction model based on BPNN and genetic algorithm to predict each component, and all the component prediction values were aggregated to obtain the ultimate predicted result. The simulation shows the purposed prediction model has higher prediction precision compare with traditional neural network prediction method and it is an effective prediction method of photovoltaic systems.

Download Full-text

Remaining Useful Life Prognosis for Turbofan Engine Using Explainable Deep Neural Networks with Dimensionality Reduction

Sensors ◽

10.3390/s20226626 ◽

2020 ◽

Vol 20 (22) ◽

pp. 6626

Author(s):

Chang Woo Hong ◽

Changmin Lee ◽

Kwangsuk Lee ◽

Min-Seung Ko ◽

Dae Eun Kim ◽

...

Keyword(s):

Deep Learning ◽

Dimensionality Reduction ◽

Short Term Memory ◽

Learning Model ◽

Remaining Useful Life ◽

Short Term ◽

Turbofan Engine ◽

Useful Life ◽

Long Short Term Memory ◽

Deep Learning Model

This study prognoses the remaining useful life of a turbofan engine using a deep learning model, which is essential for the health management of an engine. The proposed deep learning model affords a significantly improved accuracy by organizing networks with a one-dimensional convolutional neural network, long short-term memory, and bidirectional long short-term memory. In particular, this paper investigates two practical and crucial issues in applying the deep learning model for system prognosis. The first is the requirement of numerous sensors for different components, i.e., the curse of dimensionality. Second, the deep neural network cannot identify the problematic component of the turbofan engine due to its “black box” property. This study thus employs dimensionality reduction and Shapley additive explanation (SHAP) techniques. Dimensionality reduction in the model reduces the complexity and prevents overfitting, while maintaining high accuracy. SHAP analyzes and visualizes the black box to identify the sensors. The experimental results demonstrate the high accuracy and efficiency of the proposed model with dimensionality reduction and show that SHAP enhances the explainability in a conventional deep learning model for system prognosis.

Download Full-text

A Remaining Useful Life Prognosis of Turbofan Engine Using Temporal and Spatial Feature Fusion

Sensors ◽

10.3390/s21020418 ◽

2021 ◽

Vol 21 (2) ◽

pp. 418

Author(s):

Cheng Peng ◽

Yufeng Chen ◽

Qing Chen ◽

Zhaohui Tang ◽

Lingling Li ◽

...

Keyword(s):

Neural Networks ◽

Convolutional Neural Networks ◽

Prediction Accuracy ◽

Prediction Models ◽

Short Term Memory ◽

Feature Fusion ◽

Remaining Useful Life ◽

Turbofan Engine ◽

Useful Life ◽

Temporal And Spatial

The prognosis of the remaining useful life (RUL) of turbofan engine provides an important basis for predictive maintenance and remanufacturing, and plays a major role in reducing failure rate and maintenance costs. The main problem of traditional methods based on the single neural network of shallow machine learning is the RUL prognosis based on single feature extraction, and the prediction accuracy is generally not high, a method for predicting RUL based on the combination of one-dimensional convolutional neural networks with full convolutional layer (1-FCLCNN) and long short-term memory (LSTM) is proposed. In this method, LSTM and 1- FCLCNN are adopted to extract temporal and spatial features of FD001 andFD003 datasets generated by turbofan engine respectively. The fusion of these two kinds of features is for the input of the next convolutional neural networks (CNN) to obtain the target RUL. Compared with the currently popular RUL prediction models, the results show that the model proposed has higher prediction accuracy than other models in RUL prediction. The final evaluation index also shows the effectiveness and superiority of the model.

Download Full-text

A Prototype Similarity-based System for Remaining Useful Life Estimation for Future Industry by Singular Spectrum Analysis-Long Short Term Memory Neural Networks Algorithm

Journal of Mobile Multimedia ◽

10.13052/jmm1550-4646.16129 ◽

2020 ◽

Author(s):

Prakit Intachai ◽

Peerapol Yuvapoositanon

Keyword(s):

Neural Networks ◽

Short Term Memory ◽

Singular Spectrum Analysis ◽

Remaining Useful Life ◽

Window Length ◽

Short Term ◽

Turbofan Engine ◽

Term Memory ◽

Useful Life ◽

Long Short Term Memory

In this paper, we propose a prototype similarity-based approach of estimating the remaining useful life (RUL) of turbofan engine data using the singular spectrum analysis and the long-short term memory (SSA-LSTM) neural networks algorithm. The algorithm consists of two steps. First, the optimal window length of the trajectory matrix of the dataset is empirically determined from a prototype dataset. Second, the estimation of the RUL of the target datasets is performed using the window length parameter obtained from the first step. The validity of the proposed algorithm is verified by testing with 200 turbofan engine datasets. The results are shown to have a significant improvement in the performance of the RUL estimation over the existing LSTM algorithm.

Download Full-text

Online Short-Term Remaining Useful Life Prediction of Fuel Cell Vehicles Based on Cloud System

Energies ◽

10.3390/en14102806 ◽

2021 ◽

Vol 14 (10) ◽

pp. 2806

Author(s):

Tiancai Ma ◽

Jianmiao Xu ◽

Ruitao Li ◽

Naiyuan Yao ◽

Yanbo Yang

Keyword(s):

Fuel Cells ◽

Fuel Cell ◽

Prediction Model ◽

Remaining Useful Life ◽

Cloud System ◽

Commercial Application ◽

Fuel Cell Vehicles ◽

Short Term ◽

Automotive Fuel ◽

Useful Life

The durability of automotive fuel cells is one of the main factors restricting their commercial application. Therefore, establishing a remaining useful life (RUL) prediction model and developing an online operational method to apply it to the RUL optimization of fuel cell vehicles is an urgent academic problem. In this work, a short-term RUL prediction model and an online operation scheme for fuel cell vehicles are proposed. Firstly, based on historical data of a fuel cell bus under multiple conditions, the daily mode of stack voltage under a 75 A operation condition was selected as a health indicator that could better reflect the health status of a fuel cell stack. Then, an adaptive locally weighted scatterplot smoothing (LOWESS) algorithm was developed to adjust the most appropriate step size to smooth the original data automatically. Furthermore, for better prediction accuracy and stronger adaptability, a short-term RUL prediction model consisting of the adaptive LOWESS and bi-directional long short-term memory was established. Finally, an online operation scheme of the RUL prediction model based on a cloud system gave the model a strong powerful practicability. After validation, this work demonstrated good application prospects in the prognostic and health management of automotive fuel cells.

Download Full-text