Multidisciplinary Engineering Dynamic Systems Research Group
Current research highlights:
System Identification of Hydrostatic Transmission for Pendulation Control System Implementation and Simulation US Navy ship cranes are used for at-sea cargo transfer between ships for both military and humanitarian needs. Operation in rough seas can cause large payload motion (+/- 5m) in a short amount of time (30 seconds). Dr. Gordon Parker and his students are developing nonlinear control strategies for cranes that reduce these hazardous payload motions through ship motion compensation and active swing damping. Eventually, this will facilitate cargo transfer between ships while underway in extreme weather conditions.
RTR-GPS Evaluation for Crane and Payload Motion Tracking: Evaluate and implement the use of RTK based GPS for tracking the dynamic position of the payload relative to the crane tip of a shipboard based crane, using both a crane tip GPS unit as the RTK base and a stationary GPS unit on the ship’s deck. The development of payload control technology of a real-time, multi-tasking control system for payload swing suppression using sensors ranging from macro mechanical devices to MEMS rate gyros.
Graphics Hardware Accelerated Real-time Machinability Analysis of Free Form Surfaces: 70-80% of the cost of producing a product is locked in at the design stage. It is important that designers get feedback about manufacturability so that mistakes can be corrected at the design stage where the cost of correction is much lower. In the domain of Integrated Circuit (IC) design, this problem is solved by projecting manufacturing capabilities into the design through the use of Design Rule Checkers (DRCs). However, no such system exists in mechanical design. The problem of manufacturability evaluation in mechanical design is further complicated by the fact that parts are three-dimensional and there are a variety of manufacturing processes and plans that can be used to produce a single design. This project is a first cut at solving this problem. The basic idea is to find if a given free-form surface is machinable with a given set of cutters. This boils down to finding the accessibility of a cutter on a given free-form surface. Research is possible only due to the advent of powerful and flexible commodity graphics hardware called Graphics Processing Units (GPUs). GPUs are powerful parallel computing devices specially designed to manipulate large amount of graphics data for the purposes of three dimensional display. Due to the need for specialized shading methods especially in three dimensional video games, GPU vendor have enabled programmability. Researchers have used this programmability to force the GPU to perform non-graphic computations. This technique called General Purpose GPU (GPGPU) essentially hijacks the graphics pipeline using the GPU programmability for general purpose computing.
The method is essentially a GPGPU application, it works on the tessellated model of the free-form surface commonly available in all CAD systems for the purposes of three-dimensional shaded display. Instead of interpolating color for three dimensional shading, the modified graphics pipeline in this method renders interpolated vertex positions and normals to an off-screen texture. Next, a specialized fragment shader performs accessibility calculation on a per pixel basis on the projected image of the screen. The results are written to the video buffer for display.