Life Cycle Cost Analysis of Pumping System through Machine Learning and Hidden Markov Model

Abstract

The pumping system is a critical component in various industries and consumes 20% of the world’s energy demand, with 25–50% of that energy used in industrial operations. The primary goal for users of pumping systems is to minimise maintenance costs and energy consumption. Life cycle cost (LCC) analysis is a valuable tool for achieving this goal while improving energy efficiency and minimising waste. This paper aims to compare the LCC of pumping systems in both healthy and faulty conditions at different flow rates, and to determine the best AI-based machine learning algorithm for minimising costs after fault detection. The novelty of this research is that it will evaluate the performance of different machine learning algorithms, such as the hybrid model support vector machine (SVM) and the hidden Markov model (HMM), based on prediction speed, training time, and accuracy rate. The results of the study indicate that the hybrid SVM-HMM model can predict faults in the early stages more effectively than other algorithms, leading to significant reductions in energy costs.

Artificial Intelligence for Predictive Maintenance Applications: Key Components, Trustworthiness, and Future Trends

“…Refs. [70][71][72][73][74][75][76][77][78][79][80][81][82][83][84] have shown various works applying statistical and probabilistic modeling approaches such as hidden Markov models (HMMs), Bayesian networks (BNs), Gaussian mixture models (GMMs), extreme gradient boosting (XGBoost), Density-based spatial clustering (DBSC), principal component analysis (PCA), and K-means to PdM tasks. Moreover, they introduced different DNN models, such as LSTM and autoencoders, for the tasks.…”