Options
Application of Artificial Neural Network in Master Planning—A Case of Simulating the Land Use and Land Cover Changes in Bhopal
Journal
Advances in 21st Century Human Settlements
ISSN
21982546
Date Issued
2022-01-01
Author(s)
Tiwari, Vidhulekha
Chatterjee, Amit
Abstract
An urban ecosystem is intricate, and several factors collaborate to produce what we see as a city. While the dynamics of urban areas are functioning in complex ways, the conventional planning methods are generally linear and rational. This leads to an incongruity between the actual structure and the methods used to solve the resulting issues. Moreover, the accuracy and efficiency of plans prepared by conventional methods are often very less and they are not flexible. These plans do not properly respond to changes that may occur in the future. These challenges create a need to develop better planning methods for planning cities and machine learning methods can help us solve the issues. Artificial Neural Network (ANN) is one of the Machine Learning (ML) algorithms which consider various intangible, complex and nonlinear agents affecting any urban fabric. Artificial neural networks represent biological neurons and their networks. It is a black-box model, which means that the programme gets executed but the equations governing it cannot be extracted and examined. Using ANN can help in simulating various factors that are required for the preparation of a Master plan, thus improving its accuracy, time efficiency and flexibility. Using such tools can also help in better monitoring of the plans, which is otherwise time-consuming and challenging. In this study, Bhopal was selected as a case study. Even though Bhopal is the capital of the state of Madhya Pradesh, it is lagging in development as compared to several other state capitals in the country. The last master plan of the city was published in 1992, and afterwards, several attempts to make a plan have failed due to various objections arising from the side of the city’s citizens. The recent Draft Development Plan 2031, also a GIS-based Master Plan (Directorate of Town and Country Planning, Madhya Pradesh, Bhopal Development Plan‐2031, (Draft) Volume-I, Existing conditions, studies & analysis, 2020), is amid controversies regarding its ecological impacts and inaccurate mapping. In this study, ANN was used to simulate land use and land cover changes for the city of Bhopal. Using the ANN algorithm, the programme was trained with data for changes in land use and land cover from 2005 to 2018. The datasets were classified, and the algorithm was trained to check the desired accuracy. The same parameters were simulated for the year 2031 and the results were compared with the recently proposed draft development plan of the Bhopal city. The study was demonstrated how ANN could produce results that can directly be utilised as a crucial component while preparing subsequent master plans.
Subjects