P. Prempraneerach

Optimization of a z-source DC circuit breaker

DC faults may cause severe disruptions in continuity of service to vital loads in a shipboard integrated power system, hence detection, isolation, and protection against such faults must be incorporated in both medium-voltage DC (MVDC) and low-voltage DC (LVDC) systems. Here we consider the effectiveness of existing z-source breakers and propose several new designs more appropriate for fault detection in MVDC and LVDC systems. In particular, we perform an optimization study that aims to minimize dissipation and weight and we identify the key parameters for use in MVDC and LVDC systems. Preliminary verification and validation studies are also included.

Uncertainty quantification in simulations of power systems: Multi-element polynomial chaos methods

While probabilistic methods have been used extensively in simulating stationary power systems, there has not been a systematic effort in developing suitable algorithms for stochastic simulations of time-dependent and reconfiguring power systems. Here, we present several versions of polynomial chaos that lead to a very efficient approach especially in low dimensions. We consider both Galerkin and Collocation projections, and demonstrate how the multi-element decomposition of random space leads to effective resolution of stochastic discontinuous solutions. A comprehensive comparison is presented for prototype differential equations and for two electromechanical systems used in an electric ship.

Stochastic modeling of integrated power system coupled to hydrodynamics in the all-electric ship

In the paper, we model the integrated power system (IPS) of a typical destroyer ship coupled to hydrodynamics. This IPS is composed of a 21 MW, 3-phase synchronous generator supplying electric power to 4160 V AC bus and to a 19 MW, 15-phase induction motor driven by an indirect rotor field oriented control for the motor torque. Moreover, stochastic responses are examined when there exists a variation in the induction motorpsilas rotor resistance and a disturbance from the slowly-varying added resistance, derived from a Pierson and Moskowitz sea spectrum.

Delta parallel robot workspace and dynamic trajectory tracking of delta parallel robot

In a design process of a Delta or parallel robot, link parameters and actuators size must be selected to maximize appropriate workspace and payload capability. To derive a workspace of the Delta robot, the Cartesian reachable position of the Delta-robot end effector can be efficiently identified using an inverse kinematics. By varying upper-arm and fore-arm lengths, a maximize workspace for a desired reachable work area could be obtained. Furthermore, a trajectory tracking of the Delta robot for pick-and-place operation can be analyzed in terms of load-carrying ability using simplified dynamic equations such that proper actuators and gears could be correctly chosen in the design phase before constructing the Delta robot.

Integrated simulation framework for crash back operation

We present a comparison of performance of an all-electric-ship (AES) simulator using two different approaches to control the induction machine in forward as well as in cash astern operations. The first approach uses a constant-slip control while the second uses direct-torque control. We also investigate the use of proper breaking resistors required for minimal power dissipation during the crash astern operation. The AES simulator describes a medium voltage DC system (MVDC) for the USS DDG-51 Arleigh Burke-Class destroyer.

Sensitivity analysis and low-dimensional stochastic modeling of shipboard Integrated Power Systems

For design purposes, we use stochastic modeling and sensitivity analysis to identify the most important parameters of the AC power distribution in the integrated power system (IPS) model. The main AC subsystem of the IPS consists of a 59 kW synchronous generator and a motor drive of 50 hp induction motor. First, we perform sensitivity analysis treating all 31 parameters of the full model as stochastic variables. Second, we construct two low-dimensional stochastic models (with 15 and 8 parameters) and compare the multi-rate dynamics of the reduced models to the full model. For parameter variations larger than 50%, the low-dimensional models still predict accurately the mean values but underpredict the variance of most state variables.