Computational Methods for Calculating Multimodal Multiclass Traffic Network Equilibrium: Simulation Benchmark on a Large-Scale Test Case

Table 3

Solution quality and performance indicators [CT: computation time].

Indicator/method	Number of simulations	Incomplete travels (%)	Violation (%),	(euros)	Improvement to compared to GBP	CT (hours)	Improvement to CT compared to SSP	Max. number of cores used
MSA ranking	566	5,74%	13.77	1.38	−24.32%	132.26	−9.74%	1
Gap-based prob. (GBP)	594	4.08%	9.03	1.11	—	137.52	−14.10%	1
Probabilistic (prob.)	598	4.66%	11.64	1.23	−10.81%	143.19	−18.81%	1
Step-size prob. (SSP)	514	9.11%	14.77	2.61	−135.14%	120.52	—	1
Smart step-size prob.	591	5.16%	10.22	1.27	−14.41%	139.59	−15.81%	1
Simulated annealing	778	3.01%	4.82	0.32	71.17%	91.30	24.25%	3
Genetic algorithm	5866	3.79%	6.09	0.54	51.35%	83.78	30.49%	18