EFFECT OF INCORPORATING INTERSECTION DELAYS ON ROUTE ASSIGNMENTS IN AN URBAN ROAD NETWORK
T 388.13 ARD
S U M M A R Y EFFECT OF INCORPORATING INTERSECTION DELAYS ON ROUTE ASSIGNMENTS IN AN URBAN ROAD NETWORK, Emil Ardiaman, 2000, Program Magister Sistem dan Teknik Jalan Raya, Program Pascasarjana, Institut Teknologi Bandung. Generally in assignment models, route choice is modelled with link-based flow-delay curves where the travel time on a link is assumed to be function of flow on that link. However, a number of studies have shown that in urban areas, cruising speeds on most roads are virtually independent of the flows. This is not to say that total travel times are independent of flows, but the main relationships between flows and delays occur at intersections, especially on road networks with short links. The inclusion of intersection delays in route assignment models was postulated to provide a more realistic model. This study analysed the effect of inclusion of intersection delays on several inelastic route assignment models, and identified the most appropriate assignment technique for the urban area. A study was conducted in the Bandung municipal area, on a network which comprised arterial, collector, and parts of local roads, including 132 modelled intersections. The route assignment models used were: All-or-Nothing(AoN), Pure Stochastic(PS), Deterministic User Equilibrium (DUE), and Stochastic User Equilibrium (SUE). The SATURN (Simulation and Assignment of Traffic to Urban Road Network) program package was used as an analysis tool for each route assignment model. Results of the study show a significant difference in flow distribution patterns between buffer networks (modelled without intersection delay) and simulation networks (modelled with intersection delay). Average absolute differences S(g) between both networks for each assignment method are in the range of 20-30%. If the intersection delay is taken to be a part of the total travel cost routeing, then the overall increase in cost is on average 12% in DUE and SUE assignment models, and nearly 45% in PS and AoN models. The equilibrium condition with stochastic effect on the buffer network was found to be the most appropriate assumption for assignment on the Bandung network. This is shown by comparison of modelled and observed flows, where SUE in buffer network models has the lowest S(g) value (38.43%). However, this technique prevails only where travel demand (OD matrix) is used on assigned flow and is based on traffic count data without consideration of delays at intersections. iv
R I N G K A S A N PENGARUH PENYERTAAN TUNDAAN PERSIMPANGAN PADA MODEL PEMILIHAN RUTE DI JARINGAN JALAN PERKOTAAN, Emil Ardiaman, 2000, Program Magister Sistem dan Teknik Jalan Raya, Program Pascasarjana, Institut Teknologi Bandung. Umumnya pada model-model pembebanan jaringan, pemilihan cute dimodelkan berdasarkan kurva hubungan antara arus-tundaan (flow-delay curve) di ruas jalan, dimana waktu tempuh diasumsikan merupakan fungsi dari arus pada ruas tersebut. Namun, beberapa studi menunjukkan bahwa untuk daerah perkotaan, waktu tempuh di suatu ruas terlihat tidak terlalu bergantung kepada jumlah arus, akan tetapi lebih ditentukan oleh adanya persimpangan, terutama pada jaringan jalan yang pendek. Model pembebanan dengan memperhitungkan pengaruh tundaan di persimpangan, akan memberikan bentuk pemodelan yang lebih realistis. Studi ini bertujuan untuk menganalisis dampak dari tundaan di persimpangan terhadap beberapa model pembebanan serta menentukan model yang paling pantas untuk jaringan perkotaan. Studi dilakukan di kotamadya Bandung yang terdiri dari arteri, kolektor, dan beberapa jalan lokal, dengan memasukkan 132 model persimpangan. Model pembebanan inelastis yang digunakan adalah All-or-Nothing (AoN), Pure Stochastic (PS), Deterministic User Equilibrium ( DUE), dan Stochastic User Equilibrium (SUE). Sebagai alat bantu model digunakan program SATURN (Simulation and Assignment of Traffic to Urban Road Network).
Hasil studi memperlihatkan bahwa terdapat perbedaan yang nyata pada pola distribusi arus antara jaringan buffer (tanpa model tundaan persimpangan) dengan jaringan simulation (dengan model tundaan persimpangan). Rata-rata perbedaan absolut S(g) antara kedua jaringan untuk setiap model pembebanan berkisar 20-30%. Tambahan total biaya perjalanan akibat adanya tundaan persimpangan rata-rata 12% untuk model DUE dan SUE dan hampir mencapai 45% pada model PS dan AoN. Kondisi keseimbangan dengan efek stokastik merupakan asumsi yang paling pantas digunakan dalam model pembebanan untuk jaringan jalan di kota Bandung. Hal ini diperlihatkan melalui pembandingan antara arus hasil pemodelan dengan observasi lapangan, dimana model pembebanan SUE untuk jaringan buffer mempunyai nilai S(g) yang terrendah (38.43%). Namun ketentuan ini berlaku hanya jika matriks asal tujuan (OD matrix) yang digunakan, dibangun berdasarkan data arus lalu lintas tanpa memperhitungkan dampak tundaan di persimpangan. V
