HEURISTIC APPROACHES TO ROAD NETWORK OPTIMIZATION

This paper is concerned with the mathematical programming problems which arise in formulating a model to optimize the allocation of investment funds between different proposed improvement schemes in a road network. An account is first given of a number of different formulations which have been suggested and it is shown that the only acceptable models require the use of programming techniques which axe not sufficiently powerful to be able to optimize investment when more than a relatively small number of projects are considered. This has led to the development of heuristic algorithms. The remainder of the paper seeks to make the case for more careful consideration of the ways in which heuristic algorithms are employed. Although it is framed in terms of road networks, the argument is one generally applicable to many areas of engineering optimization where heuristics are used.