Dynamic O-D Demand Estimation: Application of SPSA AD-PI Method in Conjunction with Different Assignment Strategies

Table 4

Evaluation of the estimated demand as the Euclidean distance from the true O–D matrix and the total number of trips.

Reference values	Euclidean distance; seed-true; 459	Total trips; seed matrix; 12,638	Total trips; true matrix; 18,900
Experiments	Euclidean distance; seed-estimated	Euclidean distance; true-estimated	Total trips; estimated demand
Set I
Average value	826.17	820.07	17,230
St.dev.	144.04	125.74	1,714
Set II
Average value	817.15	786.86	17,786
St.dev.	92.47	83.58	2.648
Set IV
Average value	782.41	770.05	17,379
St.dev.	153.42	91.24	2,641
Set V
Average value	437.04	548.82	15,256
St.dev.	218.20	102.33	2,115
Set VI
Average value	414.52	527.81	15,183
St.dev.	199.09	80.78	1,496
Set VII
Average value	775.20	760.26	17,403
St.dev.	212.36	152.68	2,062
Set VIII
Average value	192.48	465.05	13,307
St.dev.	63.46	12.11	466