Mechanical Engineering / Makina Mühendisliği
Permanent URI for this collectionhttps://hdl.handle.net/11147/4129
Browse
4 results
Search Results
Conference Object Citation - WoS: 1Citation - Scopus: 1Modeling a Magneto-Rheological Fluid-Based Brake Via a Neural Network Method(Springer international Publishing Ag, 2022) Kucukoglu, Sefa Furkan; Dede, Mehmet Ismet Can; Ceccarelli, MarcoIdentifying the model of a magneto-rheological (MR) fluid-based brake is extremely important for designing and controlling a haptic device with hybrid actuation. Therefore, in this study, an Elman Recurrent Neural Network (ERNN) is designed to understand and model a characterization of an MR fluid-based rotational brake. Three important factors that affect the MR brake's performance are chosen as inputs: current, speed, and the first derivative of the input current. The proposed network is trained, and the performance of the network is tested with three different experimental scenarios. Then, the effect of these inputs on the system is investigated. According to the results, it can be said that the designed ERNN is a good candidate for modelling an MR brake.Book Part Citation - WoS: 1Citation - Scopus: 1Calibration Study of a Continuously Variable Transmission System Designed for Phri(Springer, 2020) Mobedi, Emir; Dede, Mehmet Ismet CanVariable stiffness actuators (VSAs) have been used in many applications of physical human-robot interfaces (pHRI). A commonly employed design is the spring-based VSA allowing the user to regulate the output force mechanically. The main design criteria of these actuation systems are the adjustment of output force independent from the output motion, and shock absorbing. In our recent work, we implemented certain modifications on the two-cone friction drive continuously variable transmission system (CVT) so that the CVT can be employed in pHRI systems. Subsequently, the optimized prototype is developed. In this study, we introduce the prototype of this new CVT systems, and its force calibration tests. The results indicate that the manufactured CVT is capable of displaying the desired output force throughout its transmission ratio range within a tolerance.Conference Object Citation - WoS: 1Citation - Scopus: 2A New Correction Coefficient Formula for the Simplified Dynamic Model of a Surgical Robot(Springer international Publishing Ag, 2021) Ayit, Orhan; Dede, Mehmet Ismet CanExecution of model-based control algorithms such as computed torque technique requires the knowledge of the dynamic model of the robotic system. In our work, the active part of the surgical robot, NeuRoboScope, has a parallel kinematics architecture and the dynamic model is relatively complicated to run in a microprocessor with limited computing capabilities. For this reason, we formulated a simplified dynamic model to run in the selected microprocessor. In this work, a new formula for calculating the correction coefficients is described to minimize the errors in the whole orientation range of the robot's base platform. This new formula is examined in terms of execution time and the result is reported in this paper.Conference Object Citation - WoS: 1Citation - Scopus: 1Alternating Error Effects on Decomposition Method in Function Generation Synthesis(Springer Verlag, 2017) Maaroof, Omar W.; Dede, Mehmet Ismet Can; Kiper, GokhanIn approximate function generation synthesis methods, error between the desired function's output and designed mechanism's output oscillate about zero error while crossing the zero error margin at precision points. The common goal of these methods is to minimize the error within the selected working region of the mechanism. For mechanisms like Bennett overconstrained six-revolute jointed linkages that have relatively large number of construction parameters, it is a difficult task to solve for them at once. Decomposition method enables to divide such linkages into two loops and independently solve for each loop with less construction parameters. Although some approximation methods are proven to produce smaller errors than others for a single-loop synthesis, in this work, it is shown that smaller errors are not guaranteed for a certain method when used along with decomposition method. Numerical examples indicate that in decomposition method, more attention should be given to the alternation of the error of each decomposed mechanism, rather than the approximation method used.