Multibody dynamics and control using machine learning

Abstract

Artificial intelligence and mechanical engineering are two mature fields of science that intersect more and more often. Computer-aided mechanical analysis tools, often multibody system software, are very versatile and have revolutionalized many industries. However, as shown by the literature presented in this review, combining the advantages of multibody system dynamics and machine learning creates new and exciting possibilities. For example, the multibody method can assist machine learning by providing synthetic data, while machine learning can provide fast and accurate subsystem models. The intersection of both approaches results in surrogate and hybrid modeling techniques, advanced control algorithms, and optimal design applications. A notable example is the development of autonomous systems for vehicles, robots, and mobile machinery. In our review we have found nontrivial, innovative, and even surprising applications of machine learning and multibody dynamics. This review focuses on applying neural networks, mainly deep learning, in connection with the multibody system method. Over one hundred and fifty papers are covered, and three main research areas are identified and introduced: data-driven modeling, model-based control and estimation, and data-driven control. The paper starts with a primer on machine learning and concludes with future research directions. The main goal is to provide a comprehensive and up-to-date review of existing literature to inspire further research.

Artificial Neural Networks With Dynamic Synapses: A Review

“…Ši tinklo struktūra buvo pasiūlyta Hochreiter ir Schmidhuber (1997) siekiant išspręsti nykstančių gradientų problemą, kuri pasireiškia įprastuosiuose rekurentiniuose neuronų tinkluose. ITAT tinklo neuronai pakeičiami atminties celėmis, kur kiekviena celė turi keturis pagrindinius komponentus: įvesties vartus, užmaršties vartus, kandidato nustatymą ir išvesties vartus (Zhang et al, 2019;Hashemi et al, 2023). Šie elementai nustato, kaip informacija patenka į celę, iš jos ir viduje.…”

Load observer for cable failure detection in Cable Driven Parallel Robots: a machine learning approach

“…To corroborate the ever growing link between multibody dynamics and artificial intelligence, several very recent studies show that combining the advantages of multibody system dynamics and machine learning is creating new possibilities and opportunities [23]. For instance, a method to replace the multibody-based equations of motion of arbitrary systems was proposed in [24] by using a deep neural network-based modeling technique, with the goal to improve computational efficiency.…”

Reliability Evaluation of Reinforcement Learning Methods for Mechanical Systems with Increasing Complexity

“…Adaptive control relies on tuning control parameters, which can be time-consuming and challenging to perform in real time. To address those challenges, research on using RL is conducted [4].…”

“…A paper by Hashemi et al [4] provides an overview of how Machine Learning might be applicable to Multibody System Dynamics. Specifically talking about RL, the authors show that its use has recently been suggested for a wide variety of tasks: starting from improving feedback control systems and robotic applications by ending with assistance in simulating human movements.…”