post thumbnail

Proposing a multi-agent deep reinforcement learning algorithm for maintenance decision-making of multi-component systems

It is widely known that maintenance decision optimization for multi-component systems faces the ‘curse of dimensionality’. In specific, the number of decision variables that need to be optimized, grows exponentially in the number of components causing computational expensive for optimization algorithms.

To address this issue, Van-Thai Nguyen, Phuc Do, Alexandre Voisin, and Benoit Iung from University de Lorraine customize a multi-agent deep reinforcement learning algorithm, namely Weighted QMIX, in the case where system states can be fully observed to obtain cost-effective policies.

Read how AI-PROFICIENT partners conducted a case study on a 13- component system to examine the effectiveness of the customized algorithm, and ultimately confirmed its performance.