Multiobjective Optimization

Robust design optimization is a modeling methodology, combined with a suite of computational tools, which is aimed to solve problems where some kind of uncertainty occurs in the data or in the model. This paper explores robust optimization complexity in the multiobjective case, describing a new approach by means of Polynomial Chaos expansions (PCE). The aim of this paper is to demonstrate that the use of PCE may help and speed up the optimization process if compared to standard approaches such as Monte Carlo and Latin Hypercube sampling.

Ships are complex and multifunctional systems. Thus their design procedure involves innumerable design variables and an overwhelming number of constraints. Under these circumstances finding an optimal solution to the design problem may seem prohibitive at a first glance. Yet there have been developed and applied many optimization methodologies over the years. A key to success is a suitable breakdown of the overall design process into manageable, well-defined modular steps, similar to the traditional decomposition of the manual design into different design stages. One of these modular steps is the definition of the watertight compartmentation. Although there are many considerations to be made for the arrangement of the internal watertight boundaries (space requirements by the equipment installed, ergonomic consideration etc), there are three that could be considered as the most critical. Those are the ship’s response in case of damage and the requirements imposed by the relevant stability regulations, the building cost of the arrangement and the transport capacity of the ship.

Given the fact that the damage stability is strongly linked to the safety level provided by the ship to its passengers, crew and cargo, the maximization of its survivability is a primary merit function for the designer. The probabilistic damaged stability regulatory concept, initially introduced by the IMO SOLAS A.265 (VIII) (1974) for passenger ships more than thirty years ago, later reaffirmed by the IMO Resolution MSC.19 (58) (1990) that was later put to force as SOLAS B-1 Reg. 25 for cargo ships built after February 1992, and lately expected to become the damaged stability regulatory standard of the future with the introduction into force of the forthcoming IMO harmonization of damage stability rules for all types of ships (see IMO-SLF, 2003) .

The present paper is originated by the relevant work of NTUA-Ship Design Laboratory within the ROROPROB EU project (see Zaraphonitis et.al., 2002 and Boulougouris et.al., 2003a) and it extends the analysis the results of the formalized multi-objective optimization procedure for the internal compartmentation of Ro-Ro Passenger ships that have been developed (see Zaraphonitis et.al., 2003 and Boulougouris, 2003b). The procedure utilized the latest advances in the field of the multi-criteria design optimization with Genetic Algorithms (GA) in order to achieve the maximization of ship’s resistance against capsize, expressed by the Attained Subdivision Index as well as her transport capacity, in terms of both increased deadweight and garage deck space. Building cost reduction is herein considered mainly as the result of steel weight minimization. Additionally it could be argued that the minimization of the number of watertight boundaries may result to equipment costs savings.

Non-recurrent events, such as lane blockage incidents or demand surge due to traffic diversion or rerouting, can increase the congestion on signalized arterial streets, resulting in long queues, queue spillbacks, and significant vehicle delays. This paper compares two methods for the selection of signal timing plans for activation during these events. First, the paper introduces a multi-objective optimization model to determine the signal timing plans considering the performance on the impacted arterial intersection approaches as well as the whole intersection performance measures during the non-recurrent events. The multi-objective optimization problem is solved via a simulation-based optimization utilizing the Non-Dominated Sorting Genetic Algorithm (NSGA-III) algorithm to find a set of Pareto optimal fronts. The Pareto optimal fronts allow trade-offs among various objectives of the optimization. Microscopic simulation models are developed for use as part of the optimization and calibrated using high-resolution controller data to better replicate real-world conditions. The performance of the resulting signal plans is compared to the performance selected using a second approach previously developed by the authors that use machine learning to emulate signal timing expert's decisions during non-recurrent events. The evaluation results show that, although both approaches can improve the performance during non-recurrent congestion, the special signal timing plans obtained from the optimization method can produce better results in improving system performance during non-recurrent events in terms of travel time, intersection delay, throughputs, and queue lengths.