Disclaimer: This is an example of a student written essay.
Click here for sample essays written by our professional writers.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UKEssays.com.

Soft Computing Based Model for Trip Production

Paper Type: Free Essay Subject: Computer Science
Wordcount: 3096 words Published: 12th Mar 2018

Reference this

Synopsis On A Soft Computing based model for Trip Production and Attraction

1. Area of the Proposed Research work

Computer Science & Applications

2. Research Topic

A Soft Computing based model for Trip Production and Attraction

3. Aim and Objectives of research work

The aim and objectives of this research is to develop a soft computing based model for trip production, trip attraction & mode-wise traffic pattern in Delhi Urban area keeping in view the development polices of the Delhi Master Plan 2021. The proposed model will greatly help in analytical study of the resulting traffic pattern and its forecasting for future city plans.

For modeling and analysis of urban transportation system, Delhi has been selected as the study zone. Data on urban activities and traffic flow was collected from the concerned agencies.

Delhi, being the Capital of India, is the main center of socio-economic, political as well as cultural activities of the country. Delhi acts as a major center of trade and commerce. The major share of travel needs of Delhi commuters is road based transport system.

4. Literature Review

This section focuses on the review of Literature for the study. It is comprises of the topics, related to general facts about the urban transport system.

The early trip generation models, based on aggregate data, predicted total trips between city pairs. The modeling methods generally include regression models, cross-classification analysis, or a combination of both. These methods still have applications due to their mathematical feasibility, data availability, and ease of interpretation (USDOT, 1999).

Kiron Chatterjee and Andrew Gordon (2006), focuses on the alternative for the Great Britain future scenarios in the year 2030 and the implications they have for transport provision and travel demand. Kiron Chatterjee and Andrew Gordon (2006) develop a National transport model to predict the national road traffics. In order to take income as a constraint for estimating trip generation there is no mathematical model under this aspect. Also, in the developing countries like India wide income disparity is there which can also plays a leading role in trip generation process.

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!

Essay Writing Service

Gravity models are signifying the idea of establishing trip distributions process. In Gravity model, the entries of the Origin-Destination matrix are understood to be a feature of the trip counts and other parameters. The main problem of the gravity model is in the measurement of travel cost. This model has been corroborated numerous times as a basic underlying aggregate association (Levinson and Kumar 1995). For the analysis zones separated by a sizeable distances, the gravity model can work properly. But, the denominator approaches to infinity as the distance between locations decreases.

In year 1947, Dantzig created the Linear Programming problem as well as offered the simplex method as its solution of Linear Programming. The simplex method was come in existence in year 1951. A simplex method is an iterative process which works along the boundaries of the problems in feasible region to find a solution of the problem. Also, the simplex method still remains the widely known solution finding technique for solving Linear Programming problems. But, the available option on the bases of Linear Programming was not successful in practice. Estimation of trip distribution is a challenging task for future period.

In order to compare the performance of intervening opportunity model, gravity model, intervening opportunity model Wilmot et.al (2006) was conducted a study. With the help of this comparison of observed trips and predicted trips, the study recommended that the traditional urban transportation planning trip distribution models are able to model trip distribution at the cumulative level, and that all models traditional models have achieved similar performance and they similar bottlenecks as well.

Anthony Chen et.al (2012) presented a basic planning tool particularly targeted at small (Metropolitan Planning Organization) MPOs was proposed to model the system traffic for the planning applications. This tool made use of PFE (Path Flow Estimator) for both the base year estimation as well as the future year predictions with some accessible field and planning data that can be available in public domains. For this tool no formal validation is done

Richter.et.al (2012) presented a model based on the logit model. The model is complex. The logit modal choice association states that the probability of selecting a mode for a trip is depend on the relative values of numerous factors such level-of-service, and travel time etc. The complex part of employing the logit modal choice model is to estimating the parameters that can be assigned for the variables presents in the utility function, so the accuracy of this logit model is not ensured.

Daniel et.al 2013, developed a inter-urban trip generation model for AkwaIbom, Nigeria, which is based on the multiple regression analysis model for forecasting future patterns. However, in a trip generation model, based on regression approach is used due to the somewhat cumbersome formulation of a choice model for frequency choice. The composite variable that would represent the service characteristics of destinations is excluded. Here, the trip generation is not based on utility maximization and the interrelationship between the trip frequencies.

5. Methodology of the research work

Following figure 5.1 shows the flow of work and the model development for trip generation (trip production and trip attraction) model presented in section 5.1, the model development for trip distribution model discussed in section 5.2 and model development for Mode-choice model discussed in section 5.3.

Figure 5.1: Flow of work.

5.1 Model Development for Trip Generation (Trip Production and Trip Attraction)

Trip generation Process widely used for forecasting travel demands. Therefore trip generation Process is divided into two parts:

  1. Trip attraction
  2. Trip Production

5.1.1. Model for Trip Production

Trip Production process focuses on the total number of trips produced from the city.

For empirical implementation of the Trip production model, Artificial Neural Network System (ANN) is used on the available data. The model was trained using data of the year 2003 & 2004 and validated on data of the year 2005. Whole process of model implementation including training and validation is accomplished in the following steps.

  1. Socio-economic data of all the zones are collected.
  2. ANN in MATLAB is used to train the model on the dataset for the year 2003 and 2004.
  3. The mode model result is validated using data set for the year 2005.
  4. Finally, trip production model is used for forecasting of the Number of passenger’s trips production for the year 2021.

5.1.2. Model for Trip attraction

Trip attraction process focuses on the total number of trips attracted by the city.

For empirical implementation of the trip attraction model, Artificial Neural Network System (ANN) is used on the available data. The model was trained using data of the year 2003 & 2004 and validated with data for the year 2005. Whole process of model implementation including training and validation is accomplished in the following steps.

  1. Socio-economic data of all the zones are collected.
  2. ANN in MATLAB is used to train the model on the dataset for the year 2003 and 2004 and then
  3. The mode model result is validated using data set for the year 2005.
  4. Finally, trip attraction model is used for forecasting of the Number of passenger’s trips attraction for the year 2021.

While validation, it was observed that the result produced by the model is very close to the actual data. The average error estimated during the validation phase is low and minimum error is only 0.8%.

5.2 Model development for Trip distribution

Trip distribution Process traveler origins and destinations to developed a “trip table” that displays the number of trips going from every origin zone to every destination zone.

For empirical implementation of the trip distribution model, Genetic Algorithm is used on the available data Whole process of model implementation including training and comparison is accomplished in the following steps.

  1. Socio-economic data of all the zones are collected.
  2. Genetic Algorithm model in MATLAB is used to implement this model on the data set to compute trip distribution for all the zones in DUA.
  3. Finally, Comparisons with of results of Linear Programming model done

Finally, comparison of Genetic Algorithm based trip distribution model with traditional linear programming is made. The result achieved from traditional Linear Programming Based Model is not up to the mark as the number of input variables increases linear programming based model gives infeasible solution.

5.3 Model development for Mode-choice model

Mode choice Process Trip distribution’s zonal interchange analysis yields a set of origin destination tables followed by; mode choice analysis allows the modeler to determine which mode of transport will be used.

For empirical implementation of the mode-choice model, Adaptive Neural Fuzzy Inference System (ANFIS) is used with surveyed data.

The model was trained using survey data for the month February 2013 and March 2013 and validated with on the survey data for the month April 2013. Whole process of model implementation including training and validation is accomplished in the following steps.

  1. Data collection from survey at different transit stations for different time periods.
  2. ANFIS toolbox in MATLAB is used to train the model for one data set of the month February 2013 and then
  3. The model parameters are modified using the second data set of the month March 2013 and
  4. Finally the mode model result is validated using data set of the month April 2013.

The model is implemented using Adaptive Neural fuzzy Inference System for peak period of work trips in Delhi urban area. The machine learning result is found quite satisfactory with validation error being as low as 0.68%.

6. Data Collection

Data on house hold population and socio-economic activities such as commercial centers, Government offices, educational institutions, and health care system were collected from Delhi Transport Corporation, Delhi Development Authority, Municipal Corporation of Delhi, C.R.R.I. (Central Road Research Institute) and also for data collection survey were conducted at transit stations where passengers have option for mode choice.

7. Implementation & Result

For empirical implementation of the trip production and attraction model Artificial Neural Network System (ANN) is used. While validation it was observed that the result produced by these models is very close to the actual data. The average error estimated during the validation trip production and attraction model phase is low and minimum error is only 0.8%.

For empirical implementation of the trip distribution model Genetic Algorithm is used. Trip Distribution model is applied on the real set of data which gives acceptable solution which is easily applicable and compared with other models such as Linear Programming model as well. The result achieved from traditional Linear Programming Based Model is not up to the mark as the number of input variables increases linear programming based model gives infeasible solution.

For empirical implementation of the proposed model Adaptive Neural Fuzzy Inference System (ANFIS) is used. While validation it was observed that the result produced by the model is very close to the survey data. The average error estimated during the validation phase is low and minimum error is only 0.68%.

8. Impact of Research

The transportation system being multidisciplinary system requires integration and co-ordination of various agencies. This includes STA (State Transport Authority), PWD (Public Works Department), DTC (Delhi Transport Corporation), DDA (Delhi Development Authority) and DMRC (Delhi Metro Rail Corporation) as well. The following table shows the impact of research in academics and industry.

Area

Impact

Academics

This research presents a new methodology for computational modeling of urban transportation system..

Industry

Traffic Planner/Economic Planner will benefit from the research methodology presented in the thesis for implementation of the urban transportation system for the region.

9. Chapter wise contents

Chapter 1 This chapter introduces the research subject giving its background and developments in the last decades. The objective of the present research is stated in this chapter.

Chapter 2 reviews the literature on different aspects of this study and presents the contribution by different researchers and deficiencies therein.

Chapter 3 presents the methodology of the research in chronological order while describing the model hypothesis and formulation in detail. This chapter also includes the description of soft computing tools used in this research.

Chapter 4 describes the formulation and empirical implementation of trip production and attraction analysis for Delhi Urban Area and computational results of this work are highlighted.

Chapter 5 describes the formulation and empirical implementation of trip distribution analysis for Delhi Urban Area and computational results of this work are highlighted.

Chapter 6 describes the formulation and empirical implementation of mode-choice behavior modelling for Delhi Urban Area .The computational results of this work are highlighted in this chapter.

Chapter 7 highlights the comments and conclusion of the research work. This also highlights the utility of this research in industry and its future scope.

10. Bibliography

[1]Rodrigue, J., Comtois, C. and Slack, B., (2006), “The Geography of Transport Systems”, Dept. of Economics & Geography, Hofstra University

[2]”Delhi Transport”. Delhi Govt. Retrieved 2014.

[3]Anthony Chen et.al, “Forecasting Network Traffic for Small Communities in Utah”, Report No. UTC-1002, February 2012

[4]Dantzing, GB, 1951, Application of the simplex method to a transportation problem, Activity Analysis of production and allocation, TC. Doopmands(ed.), N.Y. Wiley

[5]Peter Guller in SYNERGO, Planning & Project Management Zurich, Switzerland. “Integration of Transport and Land- use planning in Japan: Relevant finding from Europe”, [6]Published in Workshop on Implementing sustainable Urban Travel Policies in Japan and other Asia-Pacific Countries, Tokyo, 2-3, March 2005.

[7]Kiron Chatterjee & Andrew Gordon 2006 , Transport in Great Britain in 2030, ELSEVIER Transport Policy Journal.

[8]Levinson, David and Kumar, Ajay(1995) ‘Activity, Travel, and the Allocation of Time’, Journal of the American Planning Association

[9]FHWA. Injuries to Pedestrians and Bicyclists: An Analysis Based on Hospital Emergency

Department Data. Report No. FHWA-RD-99-078. Washington, DC: USDOT, 1999.

[10]Wilmot et.al 2006, Modeling Hurricane Evacuation Traffic: Testing the Gravity and Intervening Opportunity Models as Models of Destination choice in Hurricane Evacuation, LTRC Project No. 03-1SS, State Project No. 736-99-1116, Louisiana Transportation Research Center.

[11]Richter.et.al., Modelling Mode Choice in Passenger Transport with Integrated Hierarchical Information Integration Journal of Choice Modelling, 5(1), 2012, page no. 1-21

[12]Ekong Daniel et.al, Inter-Urban Trip Generation Models for the Urban Centers in AkwaIbom State, Nigeria, Journal of Civil and Environmental Research, Vol.3, 2013.

11. List of Publications

  • Paper Title: ‘An Improved Modeling of Mode-Choice Behavior in Urban Areas Using Adaptive Neural Fuzzy Inference System’, is present and published in the proceeding of IEEE International Conference on “Computing for Sustainable Global Development” 05-07 March 2014, New Delhi, published in IEEE Xplore.
  • Paper Title: ‘Artificial Neural Network based model for traffic production and attraction: A case study of all the zones of Delhi Urban area ’, is present and published in the proceeding of IEEE International Conference on “Computing for Sustainable Global Development” 05-07 March 2014, New Delhi, published in IEEE Xplore.
  • Paper Title: ‘Trip distribution Model for Delhi Urban Area Using Genetic Algorithm’, is published in International Journal of Computer Engineering Science March 2012; Vol.2 Issue 3, page no.1-8.
  • Paper Title: ‘A soft computing based model for traffic attraction: A case study of a segment of Delhi urban Area’ is published in Vision and Quest, Research Journal of Science Technology and Management, Issue Jan 2010, page no.42-46.
  • Paper Title: ‘Traffic Generation Model for Delhi Urban Area Using Artificial Neural Network’, is published in BVICAM International Journal of Information Technology, December 2010; Vol.2 No.2, page no.239-244 and having impact factor 0.605.

1/12

(Shivendra Goel)

Research Scholar,

Shobhit University, Meerut.

 

Cite This Work

To export a reference to this article please select a referencing stye below:

Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.

Related Services

View all

DMCA / Removal Request

If you are the original writer of this essay and no longer wish to have your work published on UKEssays.com then please: