Power System Optimisation for Naval Vessels

Research Theme: Computational Design

Optimisation of engineering systems has proven to be a useful tool in the design of complex systems, and is widely used in the aerospace and automotive industries.

This project seeks to apply these methods to the optimisation of power systems within the next generation of naval vessel.

Motivation

Due to increasing demands on power requirements in modern naval vessels, and the ever-present necessity for stealth, low operating signature and improved tactical endurance and survivability, an all-electric naval vessel is envisaged. Combined with the flexibility of replacing solid drive shafts with electrical wiring and motors and the ability for a modular design architecture, an all-electric vessel has the potential to be easier to maintain, cheaper to own and more effective at war[1].

Through the use of optimisation methods, it is hoped that the design of the drive systems within such a naval vessel will be realised.

Objectives

  • To model the drive systems of naval ships or submarines within softwarepackages such as VTB (Virtual Test Bed) and Matlab.
  • Integrate these models within the optimisation process.
  • Assess the designs practicality and select an aspect of the designto optimise.

Findings

Simulation models within VTB have successfully been integrated with optimisation algorithms such as a Multi-objective Genetic Algorithm (MOGA) and a Multi-objective Simulated Annealing (MOGA) algorithm. Visualisation tools have also been added to assist with the optimisation of the drive system. These models are currently been extended to include the drive systems within a SSBN submarine.

Details

Meta-heuristic optimisation algorithms such as MOGA and MOSA are used to optimise the design of drive systems (Fig. 2) within an all-electric naval vessel. The use of models with varying fidelity is considered as a way to decrease the computational cost while still maintaining accurate solutions.

Given the nature of drive systems, varying operating conditions must be accounted for within the optimisation process. Condition maps are considered as a way of statistically representing the range of operating conditions.

The optimisation of complex systems such as drive systems requires the simultaneous optimisation of conflicting objective functions. The optimisation process generates a set of Pareto fronts that represent the trade-offs between each objective and allows the designer to select the most appropriate solution based on the importance of each objective function (Fig.3). An effort is made to ensure these Pareto fronts are generated efficiently and with minimal cost.

Acknowledgements

Support provided by:

Support for this project was provided by the US ONR and Cambridge University Engineering Design Centre.

Selected Publications

  • [1] Hodge, C. C. G. and Mattick, C. D. J. (1998) The ElectricWarship in Transactions IMarE, vol. 108, pp. pp109-125.