Abstract

This study is concerned with assessing system performance in terms of signal timings and delay. TRANSYT, genetic algorithms (GA), SIDRA and fuzzy logic (FL) were used to obtain optimum signal settings by minimising delay. Approximate expressions were investigated for traffic measures on saturated and unsaturated approaches at signalised intersections. The sum of mean uniform and random delay together with oversaturation delay was chosen as the performance function for an intersection. The existing signal operating arrangement is also given. The basic solution algorithm for the GA is presented and applied to a sample intersection for fixed flow sets. The FL application and aaSIDRA are likewise explained in terms of solution approach and software. The results of all packages are compared in terms of signal timings, and the advantages and disadvantages of each are set out accordingly. TRANSYT produced signal timings in terms of delay. The GA achieved a better improvement than either TRANSYT or the existing arrangement. FL also outperformed the existing arrangement, whereas aaSIDRA returned delay results almost identical to it. The results offer an instructive comparison between the methods examined.