International Journal of Advanced Engineering Research and Science (IJAERS) [Vol-2, Issue-5, May- 2015] ISSN:

Transcription

1 Interpreting & Modeling Concentration in Air with Traffic Density using Artificial Neural Network with Reference to Nagpur City Shekhar L. Pandharipande 1, Shubha R. Joshi 2 1 Associate Professor, Departent of Cheical Engineering, L.I.T., RTM Nagpur University, Nagpur, India 2 B. Tech., Departent of Cheical Engineering, L.I.T., RTM Nagpur University, Nagpur, India Abstract Due to strict pollution control nors, % in exhaust of individual vehicles is decreased, however with the increase in nuber of vehicles, the total in abient air is increasing. The present work has the ain objective of interpreting and odelling the contribution of vehicular Carbon Monoxide eission in abient air. ANN odel is developed correlating the concentration of in air with the of vehicles at five road intersections of Nagpur city. Based on the observations, result and discussion, it can be concluded that the present work has successfully addressed to the proble of estiation of concentration in air with vehicular for a specific road intersection. The work is quite revealing and the situation is likely to be alaring in etro cities where traffic congestion is ore. Keywords ANN odel, eission, Nagpur city, Road intersections, Tie slots, Vehicular. I. INTRODUCTION: 1.1 OBJECTIVE: The present work has the ain objective of interpreting and odelling the contribution of vehicular Carbon Monoxide eission in abient air. Efforts have been ade to develop odel correlating the concentration of in air with the & frequency of passage of vehicles at particular road intersections of Nagpur city. The approxiate concentration of in air could be predicted at a particular place at a given tie slot when the nuber of vehicles passing through the intersection is known, using the odel developed. 1.2 BACKGROUND: Aongst the top six cities in ters of the nuber of registered otor vehicles, the highest Copound Annual Growth Rate CAGR of 13.2 % was recorded by Pune during followed by Chennai- 10.8, Hyderabad- 10.6, Bengaluru- 9.5, Delhi- 7.1 and Greater Mubai- 6.6 respectively. Other illion plus cities like Kochi, Coibatore, Madurai, Kanpur, Jaipur & Nagpur recorded ore than 10% CAGR during [1] The present study is with reference to Nagpur city, so the detailed data for year 2015 has been collected fro Sr. Category Nae of RTO Office No. Nagpur Nagpur Nagpur City East 1 2-Wheelers Motor Cars Jeeps Others Total Regional Traffic Office, Nagpur. [2] Table 1 shows the Motor vehicle population as on Of Nagpur, collected fro the RTO office. TABLE 1: Motor vehicle population in Nagpur as on The ajor part of total vehicular eission is by the petrol vehicles. It is estiated that fraction is is 1-2% of exhaust eissions for petrol vehicles, whereas it is fraction of 0.5% for diesel driven vehicles. Siilar coparison of petrol and diesel engine eissions has been given in soe works. There are soe estiates on eissions per k run for diesel and petrol driven vehicles reported in the literature. [3][4] II. METHODOLOGY The ethodology adopted in the present work is depicted in the flowchart as shown in Fig The present study is carried out at five specific road intersections of Nagpur city in the onths of winter during January & February. [5] 2. These sites have display boards ounted by local agency showing real Tie, Teperature, % huidity and concentration in pp. Page 6

2 3. Five sites at different road intersections and five tie slots of day were selected; the nuber of vehicles passing per inute was counted anually. Fig. 2: Actual photograph of the display board at Rahate Fig. 1: Flow diagra of the steps followed in conducting the present work 4. The locations of the sites chosen are as follows, Law College (L) Bhole Petrol Pup (B) Variety (V) Rahate (R) Chhatrapati (Ch) 5. The tie slots chosen were 7 a, 9 a, 1 p, 4 p, 6 p, 8 p and 11 p. 6. The eissions displayed on the boards and nuber of vehicles was tabulated and correlated using Artificial Neural Network. The vehicles were identified as 2-wheelers, 3-4 wheelers and Heavy vehicles. The counting readings were taken two to three ties for each on different working days, and the average was worked out. Fig. 2 is the photograph of the display board showing the concentration. III. OBSERVATIONS: Table 2 shows the details of vehicle at different intersections. Siilarly Table 3 gives the details of readings, tie slots and vehicle. Intersetion Ch1 Ch 2 R V 1 V 2 B TABLE 2: per inute at different intersections & tie slots Categ- tie 7 1 ory a p 9 a 4 p 6 p 10 p 2 wheelers wheelers Heavy wheelers wheelers Heavy Total wheelers wheelers Heavy Total wheelers wheelers Heavy wheelers wheelers Heavy Total wheelers wheelers Heavy Total L 2 wheelers Page 7

3 3-4 wheelers Heavy Total Fig. 3 shows the coparison of vehicle for two, three, four wheelers aongst these intersections for a typical 9 a slot V 1: Variety Square, V 2: Variety Flyover; Ch 1: Chhatrapati, Ch 2: Chhatrapati Flyover TABLE 3: readings and vehicle at various intersections Ch R Tie pp pp 7 a a p p p p Tie pp V pp B 73 Fig. 3: Coparison of vehicle at chosen sites at 9 a tie slot In Fig. 4, analysis of speed variation for a trip of 1.6 ks fro Variety to Rahate is shown. Three such trips were conducted at different tie slots of a working day and the easured paraeters are analysed to obtain the variation of speed with tie and the graph is plotted as shown. 7 a a p p p p L Tie pp 7 a a p p p p Fig 4: Variation in speed as a function of tie for travel between two intersections V to R 3.1 Graphical Interpretations: The observations as given in Table 2 are graphically interpreted by plotting the vehicle for various locations as shown in Fig. 5, 6, 7, 8 & 9. Page 8

4 Fig. 5: vehicle as a function of tie, Law Fig. 8: vehicle as a function of tie, Chhatrapati Fig. 6: vehicle as a function of tie, Bhole petrol pup Fig. 9: vehicle as a function of tie, Rahate VARIETY SQUARE conc. vs Tie slot VEHICLE DENSITY AM 9.00 AM 1.00 PM 4.00 PM 6.00 PM PM 2 wheelers 3-4 wheelers heavy vehicles pp Law sq Bhole sq Variety sq Rahate sq Fig. 7: vehicle as a function of tie, Variety Fig. 10: concentration in pp as a function of tie at chosen sites Page 9

5 Fig. 10 shows the coparison in concentration as a function of tie slots for various intersections. It can be said that, at all the intersections, concentration is lowest in the early hours, increases linearly fro 9 a to 1 p, reains alost constant during 1 p to 6 p, reaches its axiu value at 6 p and then decreases in the late evening. The concentration is lower in the early orning than late evening and night hours. IV. ANN MODEL DEVELOPMENT Artificial Neural Network Modelling of the data of concentration and vehicle has been done using elite-ann [6]. The input paraeters considered are; coded nubers for road intersections, concentration in pp for previous tie slot for respective intersection and vehicle per inute, whereas concentration for the current tie slot is the output paraeter. The training data set is shown in Table 4 used in developing ANN odel. TABLE 4: Training data set for ANN odel Coding of the road intersection concentration in pp (previous tie slot) Density per inute concentration in pp (current tie slot) Coded nubers for road intersections: 11- Ch, 22-R, 33-V, 44-B, 55-L The details of the topology of ANN odel developed is shown in Table 5. TABLE 5: Topology of ANN odel developed Nuber of neurons T D R M IP O P 1 st HL 2 nd HL 3 rd HL S P I S E , IP- Input Paraeters, OP- Output Paraeters, HL- Hidden Layer, TDSP- Training Data Set Points, I- Iterations, RMSE- Root Mean Square Error The snapshot of elite-ann in run-ode is shown in Fig. 11. Fig. 11: Snapshot of elite-ann in run-ode The coparison of the predicted concentration and actual concentration for training data set is carried out using developed ANN odel. Fig. 12 shows the bar graphics for the coparison. Page 10

6 The present work is related to study and the interpretation of concentration in abient air with vehicular by selecting five road intersections in Nagpur city. It is further extended in developent of Artificial Neural Network odel to correlate the eissions at these road intersections with vehicle. Fig. 12 : Coparison between actual and predicted values for concentration in abient air V. RESULT DISCUSSION The nuber of vehicles at Variety is axiu, and it is the busiest of the chosen sites. The axiu concentration of 4 pp is recorded at the Variety during 6 p slot of working day. On an average basis also, Variety tops the concentration followed by Law college, Bhole petrol pup, Chhatrapati and Rahate. This is supportive of the clai that concentration is linked with the eissions of nuber of petrol driven vehicles. The actual concentrations of at different locations and corresponding predicted values obtained using the ANN odel developed are very close to each other. This is indicative of the success of ANN odel developed in correlating concentration with vehicular. The odel can be further iproved with observations at ore tie slots, ore input paraeters like teperature, wind velocity, huidity, rate of dispersion etc. VI. NCLUSION Due to stringent pollution control nors, the actual eission per vehicle is decreased. However, due to urbanization, the nuber of vehicles in etro cities is increasing 10 to 15% per year, which is contributing to the increase in eissions in abient air. It is observed that, the eissions in the abient air is the function of the nuber of vehicles plying. The eission in the atosphere is the function of nuber of vehicles plying, and varies fro intersection to intersection and tie slot to tie slot. Based on the observations, result and discussion, it can be concluded that the present work has successfully addressed to the proble of estiation of concentration in air with vehicular for a specific road intersection. The work is quite revealing, and there is a need to readdress the policies regarding the nuber of vehicles plying on any road at any particular tie slot. The situation is likely to be alaring in etro cities where traffic congestion is ore. ACKNOWLEDGEMENT: The authors are thankful to RTO, Nagpur for providing the data. We are also thankful to the Director, L.I.T., Nagpur. REFERENCES [1] Road transport year book, Ministry of road transports and highways, Governent of India [2] Regional transport office, Nagpur-2015 (personal counication). [3] Narayan Iyer, A technical assessent of eission and fuel consuption reduction potential fro two and three wheelers in India, International council of clean transportation [4] R. Westerhol, Karl-Erik, Egeback, Exhaust eissions fro light & heavy duty vehicles; cheical coposition, ipact of exhaust after treatent and fuel paraeters, Environental health perspectives, vol. 102, supp , pp [5] Shubha Joshi, Project report subitted to Rashtrasant Tukadoji Maharaj Nagpur University, Nagpur for award of B. Tech. Cheical Engineering [6] S. L. Pandharipande, Y. P. Badhe, elite-ann, ROC nuber SW-1471/2004. Page 11