Research On Automatic Layout Of Building Groups Based On Multi-Agent

With the rapid development of artificial intelligence in various fields in recent years,more and more researchers have introduced artificial intelligence into the field of architecture as a method to solve complex architectural problems,and expect to use this research method to solve existing problems more quickly and efficiently,and even discuss the development trend of architecture in the information age.And this study is the cooperation projects with China united engineering co.,LTD.The project takes the industrial park as the research object,expecting to design and generate the layout of the industrial park with the help of artificial intelligence technology,and optimize the design process.In this way,architects can be liberated from the complex and boring quantitative calculation and layout trial and error,and the design efficiency can be improved.Moreover,the possibility and future development trend of industrial park architecture layout planning can be discussed through generation design.Facing such a complex problem,self-organizing multi-agent system becomes a powerful tool to solve the problem.Therefore,based on the self-organizing behavior mode of multi-agent system,the experiment applies computer programming technology to solve the principle constraint and design constraint in the planning of industrial parks and tries to construct an automatic layout model of industrial parks composed of four modules:environmental elements,building introduction,layout generation and program evaluation,exploring an automatic layout generation method of industrial parks based on multi-agent system.Therefore,this study is an attempt to solve complex architectural problems with artificial intelligence:using multi-agent system to solve the complex design problem that the industrial park layout needs to meet the requirements of multiple constraints,which replaces the boring and complicated trial and error process in the design process.Furthermore,different from the traditional optimization algorithm,the multi-agent system avoids the search of a large number of solutions,improves the efficiency of building automatic layout to a certain extent,and solves the problem of low efficiency of genetic algorithm in previous experiments.Most importantly,the model itself is not complex,although the results of the auto-layout operations reflect complex real objects.For architects who lack computer knowledge and programming skills,multi-agent systems are easier to understand and operate than other artificial intelligences.Therefore,multi-agent system is an excellent method to solve complex architectural problems.The main body of this paper is divided into four parts:the first part is the first chapter,which briefly introduces the purpose,method and significance of this research,and is the collation and summary of the industrial park functional and spatial organization characteristics and the theory of multi-agent system.The second part is the second and third chapters of the paper,as the core research results,mainly introduces the method of research,the constitution of the generation method and six constraints during the generation of industrial park layout,including sunshine spacing coefficient constraint,site off-line constraint,group aggregation constraint,height form constraint,edge rotation constraint,road square green space constraint;The third part is the fourth chapter,which verifies the automatic layout tool of the industrial park through practical cases.The fourth part is the fifth chapter,which summarizes the results and shortcomings of this experiment.At the same time,some ideas on the improvement direction of the generation tool are put forward.The whole thesis contains about 48 000 words,79 pictures and charts.