COMPARATIVE STUDY OF MLR, ANN AND ANFIS MODELS FOR ESTIMATION OF PCUS AT DIFFERENT VOLUME TO CAPACITY RATIOS
Keywords:
PCU, MLR, ANN, ANFISAbstract
PCU models are developed in the present study by taking volume-to-capacity ratios and
percentage shares of vehicle types. Field data collected on four-lane highway sections are
analysed to develop speed-flow relations. A microscopic traffic simulation model VISSIM
is also deployed for generating traffic flow data after calibrating it for mixed traffic
conditions. PCUs of different vehicle types at six lane and eight lane divided highways are
also estimated. The effect of number of lanes on PCUs was studied, and it was observed
PCU of each vehicle type decreases with increase in the number of lanes and at a
different level of service. The Adaptive neuro-fuzzy inference system (ANFIS), Artificial
neural network (ANN) and Multiple linear regression (MLR) models are used for
development of PCU Models from the results obtained through VISSIM simulation. A
comparative study was performed with PCU obtained from different models reveals that
the ANFIS model showed greater potential in predicting PCUs at varying v/c ratios and
proportional share of vehicle types in the traffic stream. The models developed in the
present study may be used for developing algorithms which describe traffic flow
behaviour under mixed traffic conditions with better accuracy and precision.
References
Aggarwal, P. (2008). Fuzzy model for estimation ofPCU. WSEAS Tran. Inform. Sci. Appli., 5(4):449-458.
Anand, S., Sekhar, S.V.C., Karim, M.R., (1999).Development of Passenger Car Unit (PCU) valuesfor Malaysia. J. East. Asia Soc. Trans. Studies3(3):73-80.
Arasan, V.T. and Arkatkar, S.S. (2010). Micro-simulationstudy of effect of volume and road width on PCUof vehicles under heterogeneous traffic. J. Trans.Eng., 136(12):1110-1119.
Arasan, V.T. and Krishnamurthy, K. (2008). Effect oftraffic volume on PCU of vehicles underheterogeneous traffic conditions. Road Trans. Res.J., 17(1):32-48.
Azamathulla, H.M., Chang, C.K., Ghani, A.A., Ariffin,J., Zakaria, N.A., and Abu Hasan, Z. (2009). AnANFIS-based approach for predicting the bed loadfor moderately sized rivers. J. Hydro-Environ.Res., 3:35–44.
Bians, M.S., Ponnu, B., and Arkatkar, S.S. (2012).
Modeling of Traffic Flow on Indian Expressways usingSimulation Technique. Procedia – Soc. Behav.Sci., 43:475-493.
Chandra, S. and Kumar, U. (2003). Effect of lane widthon capacity under Mixed Traffic Conditions inIndia. J. Trans. Eng., 129(2):155-160.
Fan, H.S.L. (1990). Passenger car equivalent for vehicleson Singapore expressways. Trans. Res.-A,24A(5):391-396.
Guisan, A., Edwards, T.C., and Hastie, T. (2002).Generalized linear and generalized additive modelsin studies of species distributions: setting thescene. Ecolo. Model., 157:89-100.
Jang, J.S.R. (1993). ANFIS: adaptive-network-basedfuzzy inference system. IEEE Trans. Syst. Man.Cybern., 23(3):665-685.
Khademi, F., Akbari, M., Jamal, S.M., and Nikoo, M.(2017). Multiple linear regression, artificial neuralnetwork, and fuzzy logic prediction of 28 dayscompressive strength of concrete. Frontiers Struc.Civil Eng., 11(1):90-99.
Khademi, F., Jamal, S.M., Deshpande, N., and Londhe,S. (2016). Predicting strength of recycledaggregate concrete using artificial neural network,adaptive neuro-fuzzy inference system andmultiple linear regression. Inter. J. Sustain. BuiltEnviron., 5(2):355-369.
Lin, C.T. and Lee, C.S.G. (1995). Neural fuzzy system.Prentice Hall, Englewood Cliffs.
Mardani, M., Ghosh, I., and Chandra, S. (2015).Determination of passenger car units on two-laneintercity highways under heterogeneous trafficconditions. J. Trans. Eng., 142(2):1-7.
Mehar, A., Chandra, S., and Velmurugan, S. (2013).Passenger car units at different levels-of-servicefor capacity analysis of multilane dividedhighways, ASCE J. Trans. Eng., 140(1):81-88.
Mehar, A., Chandra, S., and Velmurugan, S. (2014).Highway capacity through VISSIM calibrated formixed traffic conditions. KSCE J. Civil Eng.,18(2):639-645
Mehar, A., Chandra, S., and Velmurugan, S. (2016).Effect of traffic composition on capacity ofmultilane highways. KSCE J. Civil Eng.,20(5):2033-2040.
Moghaddamnia, A., GhafariGousheh, M., Piri, J., Amin,S., and Han, D. (2009). Evaporation estimationusing artificial neural networks and adaptiveneuro-fuzzy inference system technique. AdvWater Resource, 32(1):88-97.
Ramanayya, T.V. (1988). Highway capacity under mixedtraffic conditions. Traffic Eng. Control, 29(5):284-287.
Tay, J.H. and Zhang, X. (1999). Neural fuzzy modellingof anaerobic biological waste water treatmentsystems. J. Environ. Eng.–ASCE, 125:1149-1159.
Tracey, J.A., Zhu, J., and Crooks, K.R. (2011). Modelingand inference of animal movement using artificialneural networks. Environ. Ecol. Stat., 18:393-410
