Application of Long Short-Term Memory Recurrent Neural Networks Based on the BAT-MCS for Binary-State Network Approximated Time-Dependent Reliability Problems

TL;DR

This paper introduces the LSTM-BAT-MCS algorithm, combining LSTM, Monte Carlo simulation, and binary-adaption-tree methods to efficiently estimate time-dependent reliability in binary-state networks, addressing the challenge of variable component reliability.

Contribution

The paper presents a novel algorithm that integrates LSTM with Monte Carlo and BAT methods to accurately estimate dynamic network reliability, overcoming computational complexity.

Findings

Achieved at most 10^-4 mean square error in experiments

Demonstrated effectiveness on three benchmark networks

Validated the algorithm's superiority over existing methods

Abstract

Reliability is an important tool for evaluating the performance of modern networks. Currently, it is NP-hard and #P-hard to calculate the exact reliability of a binary-state network when the reliability of each component is assumed to be fixed. However, this assumption is unrealistic because the reliability of each component always varies with time. To meet this practical requirement, we propose a new algorithm called the LSTM-BAT-MCS, based on long short-term memory (LSTM), the Monte Carlo simulation (MCS), and the binary-adaption-tree algorithm (BAT). The superiority of the proposed LSTM-BAT-MCS was demonstrated by experimental results of three benchmark networks with at most 10-4 mean square error.