The Application Of Linear Programming In Parametric BIM Model

As building optimization becomes a new trend,people are using an increasing number of optimization algorithms in building design,especially the intelligent optimization algorithm,for example,genetic algorithm.They play an important role in the building optimization field.However,these algorithms are like a coin,which has two sides.Limited by the random search property,the result of multiple times of running the optimization can be different,and they can't guarantee a truly best solution obtained.On the other hand,the algorithms are packaged in the third platform,which may lead a waste of resources,and an information obstacle while BIM is also a new trend today.Considering the problems above,this paper focuses on the development,test,and application of linear programming in parametric BIM platform.First,the parametric design plugin Dynamo is introduced as the development platform,and the multiobjective optimization tool in Dynamo is presented for latter comparision.Microsoft Solver Foundation published provides the calculation engine for the program,and the related mathematical algorithms like the simplex algorithm provide the theoretical support.Additionally,the development work is conducted on the Microsoft Visual Studio 2015 IDE based on the C# language.The object-oriented property allows to pack different optimization elements in the corresponding class,which is convenient for maintenance and expansion in the future.A fast-solving pattern and a step-by-step pattern is designed aiming at different situations.The former one is designed for pure linear programming,while the latter one for a complex non-linear problem.The program is validated and explained by a mathematical example.Based on the program,three methods are used for testing the program.The first one uses randomly-generated matrix for the test,the latter two methods are used in comparision with Matlab linear programming toolbox.On one hand,the results show that the two programs have advantages and disadvantages.On the other hand,the comparision with GA-based Optimo shows that the latter one is not so stable when dealing with problems with linear constraints,and the quality of solution is worse than the developed method.Finally,a solar heat gain optimization is conducted,aiming to showcase the application of linear programming in building performance optimization,and the steps when applied.It is worth noting that the application of linear programming requires an appropriate and specific mathematical model,which may be a limitation of the program.