Abstract
The design of many engineering systems, including sonar systems, has evolved (thanks in part to methods of operational research such as mathematical programming and simulation) from an almost strictly art form into a relatively scientific endeavor. Today we see, more and more, the use of, in particular, mathematical programming methods for the development of ‘optimally’ designed engineering hardware. Unfortunately, such designs are ‘optimal’ only under the unrealistic assumption of an error free model and the ability to develop a corresponding perfect design. That is, they fail to take into regard the often severe impact of the limitations of real world design tolerances. And this failure is becoming increasing more important as we begin to address modern day systems with their substantially increased size, sophistication and sensitivity. In this paper, we discuss the problem, with specific focus on the task of the development of sonar arrays. We present a hybrid methodology, combining expert systems with heuristic programming, that permits the engineer to incorporate the results of the mathematical programming solution while still using real world, imperfect components. And we present results as obtained on actual hardware.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call