ITcon Vol. 16, pg. 357-380, http://www.itcon.org/2011/22

Procedural lot generation for evolutionary urban layout optimization in urban regeneration decision support

published:February 2011
editor(s):Turk Z.
authors:Syed A. Yusuf, Dr.,
School of Engineering and the Built Environment, University of Wolverhampton, Wolverhampton, United Kingdom;
syed.yusuf@wlv.ac.uk;http://www.wlv.ac.uk/default.aspx?page=17592

Panagiotis Georgakis, Dr.,
School of Engineering and the Built Environment, University of Wolverhampton, Wolverhampton, United Kingdom;
p.georgakis@wlv.ac.uk

Christopher Nwagboso, Prof.,
School of Engineering and the Built Environment, University of Wolverhampton, Wolverhampton, United Kingdom;
c.nwagboso@wlv.ac.uk
summary:Decision support in urban regeneration planning involves fulfilment of complex design objectives that require a close collaboration between stakeholders, designers and planners. The objectives are to achieve computationally optimized design plans that are environmentally robust and sustainable. For the purpose of pre-planning analysis and optimization of such design models, the construction framework of the whole system development life cycle is simulated using specialist analysis tools. In the analysis and design phase of such a construction process, location allocation of built environment structures integrated with AI based urban assessment models remain a cumbersome task. With the advent of sophisticated computing hardware and 3D modelling techniques, it has become viable to visualize and develop simulation outcomes with ease over standard computers. However, creation of simulated outcomes with huge urban details still remains a daunting task. With the advent of knowledge-based data standards and mining techniques, possibilities have arisen to integrate procedural urban modelling frameworks with socio-economic deprivation assessment systems to create massive city models in order to improve sustainability and smart growth. The implementation of a procedural modelling framework offers a promising area of research in location allocation optimization of residential and public service structures for urban regeneration planning and collaboration purposes but is still largely stochastic in nature until now.The work presented in this article proposes an L-system-based automated urban layout generation module in addition to an evolutionary building placement optimization system for the purpose of urban regeneration decision support. The location allocation optimization of a highly accessible pedestrian and street network grid still remains a complex process especially with neighbourhoods suffering from severe socio-economic deprivation. Using an extension to ‘L-systems’, the module utilizes an ‘mBPMOL’ system to recursively subdivide an urban regeneration layout and later encode the layout to a genetic chromosome to iteratively place a range of urban structures to the lots. Moving further, the system implements a distance-based socio-economic deprivation minimization fitness function and assesses subsequent individual genetic solutions to search for an optimized layout.The development of such an online procedural framework is a very first attempt to employ the capabilities of procedural automation with evolutionary computation to automatically generate and optimize the placement of a large number of building models. The outcome achieved a set of relatively optimal solutions with evaluation based upon a distance-based fitness objective function of various public service structures. The solutions (layout plans) thus obtained offered a number of moderate to highly accessible alternatives. Development of such a system is meant to facilitate decision support and improve time efficiency among urban designers and planners.
keywords:Artificial Intelligence, Genetic Algorithms, Urban Regeneration, Built Environment, Urban Planning
full text: (PDF file, 1.012 MB)
citation:Yusuf S A, Georgakis P, Nwagboso C (2011). Procedural lot generation for evolutionary urban layout optimization in urban regeneration decision support, ITcon Vol. 16, pg. 357-380, https://www.itcon.org/2011/22