Table 3.

Performance analysis of LTCOP-LSTM model for traffic flow prediction in different time periods.

Time interval (Hours)	Traffic flow (vehicles/h )	Model latency (s)	Error fluctuation (vehicles)	Processing efficiency (%)	Dynamic modeling time (s)	Feature engineering complexity	Number of diverse input features
0–1	1500	12.34	6.78	92.5	8.45	High	5
1–2	1800	15.22	7.15	91.0	10.37	Medium	6
2–3	1600	14.85	5.90	93.4	9.01	Low	4
3–4	2000	17.46	8.56	89.0	11.65	Medium	7
4–5	2200	16.78	9.22	88.5	10.12	High	8
5–6	2400	19.38	10.03	87.5	13.44	Medium	5
6–7	1950	13.55	7.75	90.4	8.88	Low	6
7–8	1750	15.90	8.12	91.6	9.67	High	7