In this investigation, the nonlinear swing-up problem associated with the cart-pole system modeled as a multibody dynamical system is solved by developing a deep Reinforcement Learning (RL) controller. Furthermore, the sensitivity analysis of the deep RL controller applied to the cart-pole swing-up problem is carried out. To this end, the influence of modifying the physical properties of the system and the presence of dry friction forces are analyzed employing the cumulative reward during the task. Extreme limits for the modifications of the parameters are determined to prove that the neural network architecture employed in this work features enough learning capability to handle the task under modifications as high as 90% on the pendulum mass, as well as a 100% increment on the cart mass. As expected, the presence of dry friction greatly affects the performance of the controller. However, a post-training of the agent in the modified environment takes only thirty-nine episodes to find the optimal control policy, resulting in a promising path for further developments of robust controllers.

In this investigation, a closed-chain kinematic model for two-wheeled vehicles is devised. The kinematic model developed in this work is general and, therefore, it is suitable for describing the complex geometry of the motion of both bicycles and motorcycles. Since the proposed kinematic model is systematically developed in the paper by employing a sound multibody system approach, which is grounded on the use of a straightforward closed-chain kinematic description, it allows for readily evaluating the effectiveness of two alternative methods to formulate the wheel-road contact constraints. The methods employed for this purpose are a technique based on the geometry of the vector cross-product and a strategy based on a simple surface parameterization of the front wheel. To this end, considering a kinematically driven vehicle system, a comparative analysis is performed to analyze the geometry of the contact between the front wheel of the vehicle and the ground, which represents a fundamental problem in the study of the motion of two-wheeled vehicles in general. Subsequently, an exhaustive and extensive numerical analysis, based on the systematic multibody approach mentioned before, is carried out in this work to study the system kinematics in detail. Furthermore, the orientation of the front assembly, which includes the frontal fork, the handlebars, and the front wheel in a seamless subsystem, is implicitly formulated through the definition of three successive rotations, and this approach is used to propose an explicit formulation of its inherent set of Euler angles. In general, the numerical results developed in the present work compare favorably with those found in the literature about vehicle kinematics and contact geometry.

No abstract

No abstract

In this study, a novel technique for multiple damage detection of structures using modal characterization to evaluate the dynamic response of the structure given a damage model is investigated. The damage identification problem is seen as an optimization problem to be solved using a firefly optimization algorithm. The objective function is based on a numerical damage model that considers the modal response of the structures. We show some implementation details and discuss the obtained results for a benchmark problem used to assess the performance of the method and its advantages for structural health monitoring