This paper describes the simulation and control of a PC-based driving simulator that is constructed by five hydraulic cylinders and can perform five degree-of-freedom(DOF) motions. The mathematical equations of vehicle dynamics are derived from the 2-DOF bicycle model and incorporated with the tire, steering and suspension sub-systems. These equations of motion are then programmed by MATLAB, transferred into C++ code in MIDEVA environment, and further developed into motion platform control program by C++ Builder. In addition, virtual reality environment is projected onto a 72" screen to present the street scene and simulate the sixth DOF motion of a vehicle, and can be employed either in off-line analysis and simulations or on-line simulations. In the future, a road surface model and other more complicated vehicle dynamics models will be incorporated, and a pressure vest will be installed to stimulate the driver to feel the longitudinal translation.

View full abstract

Two nonlinear sliding observers are developed for speed-sensorless control of permanent magnet synchronous motors. One of the observers estimates the rotor position and velocity based on the measurement of the motor current, while the other estimates load torque as well as rotor position and velocity. The lower bounds of the sliding gain, which guarantee the existence of the sliding manifold, and the upper bounds, which guarantee the stability of the digital implementation of the sliding observers, are also presented. Both theoretical analyses and experimental results demonstrate that, regardless of the observer used for speed-sensorless control, the steady-state velocity error converges even when an external load torque is applied to the rotor. However, when an external load is present, the position estimation error of the control system with torque estimation and compensation is significantly smaller than that of the system without torque estimation and compensation.

View full abstract

This paper presents a simplistic binocular robotic head architecture and a thorough analysis of its rigid body dynamics and kinematic models. A kinematic mathematical model for online computations of desired joint angle trajectories in object centroid tracking and fixation is presented. The efficacy and the characteristics of the models derived are illustrated using dynamic computer simulations with a controller using computed torque method.

View full abstract

In this paper, a three-dimensional parallel solution scheme for inverse dynamics of link mechanisms is described, which is already proposed in a two-dimensional case and applied in several in-plane motions. In this theory, the entire system is subdivided into finite elements and evaluated as a continuum. A single-link structure of a pin joint and a rigid bar is expressed by using the shifted integration (SI) technique, which is usually used in finite element analyses of framed structures. This scheme calculates nodal forces by evaluating equations of motion in a matrix form, and thus information from the entire system can be handled in parallel. The obtained nodal forces are then converted to the joint torque in the system. Simple numerical tests on two-dimensional and three-dimensional open-loop link mechanisms are carried out to compare them with other schemes, and some tests on closed-loop and continuously transforming mechanisms are carried out to confirm the flexibility of the scheme.

View full abstract

This paper presents thermal and flow analyses and optimization of a bi-directional valve-less micropump. The bi-directional pump does not use mechanical check valves but takes advantage of the viscosity variation according to temperature. When there is a viscosity difference between two outlet chokes, the flow rate of one choke becomes greater than that of the other, which results in flow rectification in one direction. The benefits of this type of micropump are: its structure is simple, and it allows bi-directional pumping. In order to investigate the behavior of the micropump, an FEA model is made for the whole physical domains, and coupled thermal and fluid flow analyses are conducted using a commercial code throughout the entire working cycle. The rectification efficiency and fluid flow rate are designated as the objective function for optimization. To improve the performance of the pump, important thermal parameters-heat flux, preheating time and heating time-are selected and optimized.

View full abstract

Researches have shown the cutting parameters and the burrs formation are close related, but these relationships cannot be formed in a simple formula. The mechanism of burrs formation and its appearance occurred to the cutting material quite depends on the method of machining and cutting condition. Although the relationships are nonlinear, the residual burrs can be reduced significantly by selecting appropriate cutting parameters during machining. In this research, a series of cutting experiments that based on Taguchi experimental method has been conducted to explore the formation of burrs size and types under different cutting conditions. The relationship of cutting parameters and the burrs formation data are collected for further study. With the burr size as the evaluation index, cutting speed, feed-rate, and depth of cut are chosen to cut the medium carbon steel (S50C). The Taguchi Method and Artificial Neural Network are adapted to establish the burrs formation model, and then the neural network based on optimal design method as a tool in cutting parameters optimization is employed. The result shows the goal of reduce burrs size into a reasonable region can be accomplished by adjusting cutting parameters. The experiments proved the burrs size with the optimal design method can be reduced as much as 67 to 78% that comparing with experienced cutting condition. As this point of view, the parameters optimization operations by optimal parameters design method offer an effective tool to reduce the burrs size in machining.

View full abstract