Model Split Analysis for the Selected Corridor- Bangalore City

IJSRD - International Journal for Scientific Research & Development| Vol. 3, Issue 11, 2016 | ISSN (online): 2321-0613 Model Split Analysis for the Selected Corridor-Bangalore City Lohit Talawar Assistant Professor Department of Civil Engineering VSMIT Nippani, Belagavi, Karnataka, India Abstract— A study corridor of length 16.5km was selected for the study. This stretch mixed type of land uses i.e. residential, commercial and recreational. Data on travel time by bus and car, opinion surveys on fare, mode changes, comfort and LOS were conducted. Along with them trip purpose, travel time and waiting time at stops were also used. Analysis of the data indicated large percentage of trip performed with work purpose followed by education, maximum number of commuters travel to a distance of more than 5km. Time spent in waiting for the bus was also a higher side and about 56% commuters were satisfied with the present LOS. About 78% of commuters were ready to shift to the higher modes of transportation. An utility function was derived to express the satisfaction experienced by the commuter. This is significant at 5% level of significance of f-test. Key words: Transportation, Model Split Analysis I. INTRODUCTION Transportation is a vital component of economic development, social progress and quality of life of urban transportation. The number of trips made by different modes within and outside the area depends on its accessibility to the required services and facilities. Cities play a vital role in promoting economic growth and prosperity of a nation. The development of cities largely depends upon their physical, social, and institutional infrastructure. Transport demand in most Indian cities has increased substantially, due to increases in population as a result of both natural increase and migration from rural areas and smaller towns. Increases in household income, and increases in commercial and industrial activities have further added to transport demand. In many cases, demand has outstripped road capacity. Greater congestion and delays are prevalent in Indian cities and indicate the seriousness of transport problems. A high level of pollution is another undesirable feature of overloaded streets. The transport crisis also takes a human toll. Statistics indicate that accidents are a primary cause of accidental deaths in Indian cities. The main reasons for these problems are the prevailing imbalance in modal split, inadequate transport infrastructure, and its suboptimal use. Public transport systems have not been able to keep pace with the rapid and substantial increases in travel demand over the past few decades. Bus services in particular have deteriorated, and their relative output has been further reduced as passengers have turned to personalized modes and intermediate public transport. Modal split is the process of separating person’s trips by the mode of travel. In general, modal split refers to the trips made by private car as opposed to public transport with respect to passenger modal split. Modal split continues to receive primary attention from investigators interested in urban transportation modeling, probably because completely acceptable models have yet to be developed and because the modal split problem is of the most concern while deciding among alternative transportation proposals. In recent years one of the most favored approaches to modal split has been to pose problem as one of general consumer choice and utilize logit analysis to solve it(2). A. Objectives of the Dissertation The main objectives of present study are:  To study the travel attributes of the commuters in the selected corridor.  To conduct opinion survey in the selected corridor and to perform modal split analysis.  To develop utility function for selected corridor. II. FIELD SURVEY AND DATA COLLECTION A. General 1) Study Area Details: The project road stretch selected for the present study is from Majestic Bus Station also known as the Kempegowda Bus Station (KBS) near Gandhi Nagar which is quite neatly organized and easily navigable to Yelahanka Satellite Town also known as New Yelahanka. The total stretch length was 16.5 kms from Kempegowda Bus Station to Yelahanka Satellite Station and 17.5 kms from Yelahanka Satellite Station to Kempegowda Bus Station. Based on the interview with BMTC authority principal-HRD, central offices, shantinagara and chief Traffic managers, divisional controllers, Depot Managers of BMTC organization, the route 402-D was selected because, the corridor between CBD and areas connecting devanahalli has been seeing major investments after establishment of new International Airport. The major bus stops along the study area were,  Kempegowda bus station  Shivananda circle  Gutahalli  Mekhri Circle  Hebbal  Military dairy farm  Byatarayanapura  Allalsandra  Yelhanka NES office Yelahanka satellite station Fig. 1: Map showing Study Area along KBS and YST All rights reserved by www.ijsrd.com 238 Model Split Analysis for the Selected Corridor-Bangalore City (IJSRD/Vol. 3/Issue 11/2016/054) B. Data Collection Selection of a test stretch: KBS to Yelahanka Satellite Station Number of sample: 1000 Number of Enumerators: 3 Conducting Opinion survey at five stations namely KBS, Hebbal, CBI, NES and YST. Opinion survey content of trip purpose, trip distance, waiting time, level of service, and commuter opinion about shifting to higher mode of transport like metro and mono rail. Bangalore Metropolitan Transport Corporation (BMTC) is the only running the bus satisfaction to travel demand of the people 1) In-Bus Survey The survey was conducted by in-bus survey for the route number 402-D for one week. The surveys were conducted from 8.00 am to 8.00 pm thus covering both peak and offpeak periods in both directions of travel. Number of enumerators involved are 2. The survey numbers of trips surveyed are given in the table below Route Number Number of trips 402-D 14 Table 1: In-bus Survey Details Table 1 In-bus survey details. The survey was conducted in 402-D buses for one week. 14 trips in-bus survey was carried out. The average journey time taken from origin KBS to destination YST is 52 minutes and from origin YST to destination KBS is 62 minutes. It is to be noted that city buses have two doors-one in the front and the other in the rear. Boarding and alighting is allowed in both the doorways as ladies and gents are allowed to use only front and rare doors respectively. The following data were collected from IN-BUS survey  Stopped time at each bus stop with causes i.e., only for B & A or with delay due to issuing of tickets also.  Other delays caused enroute with causes and duration.  Time of reaching each of the pre-specified points along the route.  Starting and ending time of the trip. 2) Opinion Survey Enumerators were trained and used to collect data from various locations along the study route by interviewing trip makers. Ten stage points were identified for conducting interviews of the trip makers. Primary data includes the Socio-economic information such as age, gender, etc of the respondents. Trip Characteristics such as OriginDestination, purpose of the trip, fare paid, and time spent in the bus stops etc. Modal choice details such as Metro, Monorail, Fare willing to pay for different modes were interviewed. The interview was conducted for seven days. The survey was conducted by 3 Enumerators on all seven days along study area. Totally 1000 (one thousand) samples were collected. 3) Car Travel Time Survey Car survey was conducted in the study area along the bus route. The enumerators were asked to note down the journey time between Kempegowda Bus Station to Yelahanka satellite station. Car survey was planned to conduct during both peak and off-peak hours. An average journey time was then found out. Delay time due to fixed signal points and delays due to other causes (operational delays) were also noted down. The average journey time taken from origin KBS to destination YST is 38 minutes and from origin YST to destination KBS is 43 minutes III. ANALYSIS OF THE DATA The analysis of the data can be categorized into three parts:  Opinion survey Analysis  Analysis of Modal Split  Utility Function A. Opinion Survey Analysis: In this step all 7 days data was combined together for study area. Bar charts and pie charts are plotted for Trip purpose, Trip distance, Waiting time at bus stops, Level of service, Willingness to shift to metro ,Willingness to pay to metro, Willingness to shift to monorail and Willingness to pay to monorail respectively .The graphs have been shown below figures. 1) Trip Purpose The data analysis result indicates 41% of commuters perform education trips, 35% commuters perform work trips, 13% commuters perform business trips, 7% commuters perform social trips and 4% of other commuters Figure-2. Shows the variation of trip purpose in the form of pie chart Fig. 2: Pie chart showing Percentage showing the various trip purposes 2) Trip Distance The total distance of the stretch is 16.5 kilometers. The analysis shows 69.16% commuters travel more than 5 Kilometers distance. 23.4% commuters travel distance between 3 and 5 Kilometers and 7.38% commuters travel distance less than 3 Kilometers. Figure-3 indicates the graphical representation of trip distance Fig. 3: Graphical Representation of trip distance 3) Waiting Time Form the Figure-4, 71.68% commuters wait for bus at bus stops for around 6 to 15 minutes, 17.26% commuters wait for more than 15 minutes and 11.07% commuters wait for less than 5minutes. During survey it has been noticed that the commuters travelling for shorter distances need not wait All rights reserved by www.ijsrd.com 239 Model Split Analysis for the Selected Corridor-Bangalore City (IJSRD/Vol. 3/Issue 11/2016/054) for more time, as this stretch covers more than half a distance in common with other routes passing this way. Fig. 4: Graphical Representation of Waiting Time 4) Level of Service Based on the opinion survey, 38.92% commuter’s felt the present LOS is good, 56.48% commuter’s opinion was fair and only 4.59% commuter’s opinion was poor. This indicates that the service of BMTC is good with respect to frequency of buses and number of passengers, availability of seats, behavior of conductors with commuters, bus condition etc, as indicated in Figure-5. Fig. 5: Graphical representation of level of service 5) Willingness to Shift to Metro Rail The survey result indicates 78% commuters prefer to travel in metro rail and 22% commuters prefer to travel in BM TC only and refuse to travel in metro rail. The survey data is presented in the form of pie-chart in Figure-6. Fig. 6: Pie Chart showing percentage of passenger’s willing to shift from BMTC to Metro 6) Willingness to pay for metrorail The minimum Metrorail fair is Rs.10 from one stop to another, and maximum is Rs.15 in the completed Metrorail stretch from Baiyappanahalli to M G Road. Based on these fair, survey was conducted which indicates 7.43% commuters prefer to pay the present fair of Rs.10, 28.2% commuters are willing to pay Rs.13 and 64.35% commuters are willing to pay Rs.15. Figure-7 indicates the percent of commuters willing to pay for metro. Fig. 7: Relationship showing commuter V/s willingness to pay for metro Figure-7: Relationship showing commuter V/s willingness to pay for metro Fig. 8: Graphical showing percentage of passenger’s willingness to shift from BMTC to Monorail 7) Willingness To Pay For Monorail Figure-9 indicates 25.47% commuters prefer to pay the fair of Rs.10, 70% commuters are willing to pay Rs.15 and 4.53% commuters are willing to pay Rs.5 based on the survey results. Fig. 9: Graph shows percentage of passenger’s willingness to pay for Mono Rail B. Analysis of Modal Split 1) Mode Choice Models Contemporary mode choice (or modal split) models are almost always disaggregates models based on a utility function of the form �� = � + � � ��� + � Where, Um = Utility of mode m, βm = Mode-specific parameter, zmj = Set of travel characteristics of mode m, such as time or money, αj = Parameters of model, to be determined b calibration. δ = Stochastic term with zero mean. The term βm expresses the relative desirability of different modes to the members of the market segment, and is assumed to capture the effect of all characteristics of the mode that are not included in the Z terms. The stochastic term δ express the variability in the individual utilities around the average utility for members of the market All rights reserved by www.ijsrd.com 240 Model Split Analysis for the Selected Corridor-Bangalore City (IJSRD/Vol. 3/Issue 11/2016/054) segment. It represents some kind of probability distribution function, and is assumed to have zero mean, so that the rest of the utility function gives the average utility of the members of the market segment for mode m. other distributional properties of δ depend on the type of distribution assumed, as does the form of the resulting demand model. 2) Multinominal Logit Model The multinomial logit model demand equation is given by ��� �� = ∑� � � Where, Pm = Probability that mode m is chosen e = base of natural logarithms. m = index over all modes included in the choice set C. Utility Function Making the urban traffic mode structure change to an optimal state, it is necessary and important to establish applicable utility function of the urban passenger traffic mode selection. The main influencing factors are chosen from the urban resident trip survey data, and will be used as utility function parameters. To correct the defects of the traditional linear weighted utility function, the utility function without inter-factor correlations is expressed by the multi-dimensional vectors, based of which, the traffic modesplit model is established. Then the distance between vectors is defined with the inner product method and the vectors are sorted with the distance analysis method. The calculation example in the last shows that, the new utility function can make the traffic mode-split model better reflect the relationship between the travel distance and the mode-split rate. Because the mode-split value is more reasonable and better fit the real case of travel mode selection. better fit the real case of travel mode selection. 1) Development of Utility Function Utility function expresses the consumer’s indifference between various alternative choices or attributes of these choices. The generalized Utility equation is given by U = βm – α1C – α2T............................. (3) Where, βm = calibrated mode specific constant. α1 and α2 are constants, C = Out of pocket costs, T = Travel Time, sec Mathematically Utility function is given by U = f (x, y) …………. (3.1) The above equation can be re-written as U= f (Cost, Travel time, Comfort)…… (3.2) The coefficients associated which each variable has been derived as shown. Busfare=1.10+2.30*Journeytime+1.10*Comfort- 0.10*LOS The above function of utility has been used to compute Modal Split. Taking bus fare as dependent variable, journey time, comfort and level of service (LOS) as dependent variables. After running the program for utility function, following utility model is obtained; Table-2 shows the statistics of the utility function Parameter Value Sample size 5 Adjusted R2 0.887 p 0.032 F- value 11.417 Table 2: Statistics of the Utility function Since the p-value is less than 5%, it states that the sample fairly represents the population. The strength of the model indicated by its R2 value which is about 0.887. Considering the F-value, indicates the model is significant at 5% level of confidence. IV. CONCLUSION Following conclusion could be made from Analysis 1) The average bus journey time taken from origin Kempegowda Bus Station to destination Yelahanka Satellite Station is 52 minutes and from origin Yelahanka Satellite Station to destination Kempegowda Bus Station is 62 minutes. 2) The average car journey time taken from origin Kempegowda Bus Station to destination Yelahanka Satellite Station is 38 minutes and from origin Yelahanka Satellite Station to destination Kempegowda Bus Station is 43 minutes. 3) The selected stretch entertains maximum percent of work trips followed by education trips. 4) Maximum number of passenger felt satisfaction with respect to waiting time at the bus stations. 5) The waiting time and level of service offered by the present transit system found to be adequate for the present sample of commuters. 6) Maximum percentage of commuters travels up to 5 kilometers. 7) About 70 percentage of commuter’s show inclination to shift to Metro rail from present system 8) Maximum percentage of commuters of bus transit is willing to pay current fares of Metro as prevailing in Bangalore 9) Less number of commuters prefers Mono rail compared to Metro rail, since they are not aware of mono rail system. 10) As a measure of performance of the bus system, an utility function is derived. This could be used along with the choice models to measure benefits quantitatively. REFERENCES [1] http://en.wikipedia.org/wiki/Talk:Bangalore. [2] http://nptel.ac.in/ [3] Simha N R N, Veeraraghavan and R Sathyamurthy, “Optimisation of Bus Route Network of Bangalore City-A Case Study”, Bangalore, 1992. [4] Bangalore Citizen Perceptions on Democratic Capital, 2011. [5] Lelitha Devi Vanajakshi, “Estimation and Prediction of travel time from loop detectordata for intelligent transportation systems”, University of Kerala, India, 2004. [6] Van Lint J.W.C, “Reliable Travel Time Prediction for Freeways”,Netherlands, 2004. [7] Todd Litman, “Transportation Cost and Benefit Analysis II – Travel Time Costs Victoria Transport Policy Institute”, Victoria, Canada, 2007. [8] http://vtpi.org/ All rights reserved by www.ijsrd.com 241 Model Split Analysis for the Selected Corridor-Bangalore City (IJSRD/Vol. 3/Issue 11/2016/054) [9] Ashalatha R; Manju V. S; and Arun Baby Zacharia, “Mode Choice Behavior of Commuters in Thiruvananthapuram City” NATIONAL INSTITUTE TECHNOLOGY - SURATKAL on 11/12/13. [10] Sathish. H. S, Jagadeesh. H.S, Skanda Kumar, “Travel Delay and Modal Split Analysis – A Case Study” IOSR Journal of Mechanical and Civil Engineering (IOSRJMCE), Jan. - Feb. 2013. [11] John Theologitis and David Powell, “Model Split in Recreational Transport Planning” Journal of Transport Engineering, February 1984. [12] Rastogi Rajat and Krishna Rao. K. V, “Travel Characteristics of Commuter Accessing Transit” Journal of Transport Engineering, December 2003. [13] Tom V. Mathew and Krishna Rao. K. V, “Introduction to Transportation Engineering” NPTEL May 7, 2007. [14] Buehler, R. “Determinants of transport mode choice: A comparison of Germany and the USA,” 2011. [15] Liu, G. “A behavioral model of work-trip mode choice in Shanghai.” 2007. [16] Nurdeen, A., Rahmat, R. A. O. K., and Ismail, A. “Modeling of transportation behavior for coercive measures for car driving in Kuala Lumpur.” ARPN J. Eng. Appl. Science, 2007. [17] Narasimha Murlhy. A. S and Ashtakala. B, “Modal split analysis using logit models” Downloaded from ascelibrary.org on 11/12/13. All rights reserved by www.ijsrd.com 242