Automation of electro-hydraulic routing design using hybrid artificially-Intelligent techniques

Abstract

Traditional ‘simple’ genetic algorithms are theoretically capable of solving the 3-D spatial problem represented by hydraulic and electrical harness design. However, the size of the ‘solution space’ to be searched, for even the simplest of problems can represent a computational load sufficient to limit any practical application. This research proposes a ‘key-point search’ which when used prior to the GA can successfully reduce the size of the computational task. It does this by identifying those points in the physical three dimensional space which are most likely to be useful in the final solution and producing an initial population of solutions from these points. This is shown to significantly reduce computation times to find valid solutions.