Research On The Efficacy Of Optimization Algorithms Used In Building Energy Optimization

In recent years,performance-oriented architectural design and optimization methods,which plays an important role in promoting green,low-carbon,energy efficient architectural design,has become a hot research direction of international architecture.Energy consumption is one of the most important,most common and most concerned performances of buildings.Therefore,building energy optimization(BEO)has become the main research topic in the field of performance-oriented building design and optimization methods.The optimization algorithm is an important part of the BEO technique,and has a decisive impact on the effectiveness and efficiency of the technique.It plays the same key role in BEO as the aero-engines in aircraft.However,the current research on optimization algorithms in BEO is still insufficient.There lacks scientific understandings of when the algorithm is effective,when it fails,why it fails,and how to choose proper algorithms when faced with different BEO problems.Besides,the guidance on selecting proper algorithms based on the properties of BEO problems is still inadequate,which leads to significant difficulties in design practice,such as inefficiency and inability to obtain optimal design solutions.Through the intersection of architecture,computer science,mathematics and other disciplines,this paper systematically carried out the research on optimization algorithms in BEO.The main contents of the thesis are as follows:(1)Based on the general optimization theory,the properties of BEO problems are analyzed from the perspective of multiple parameters.Two methods for determining the properties are proposed,including numerical method and analytical method.(2)A set of indices is proposed to evaluate the overall efficacy of optimization algorithms in BEO.For each index,the mathematical definition and the quantitative calculation method are provided.(3)The ineffectiveness of optimization algorithms in BEO is defined.Five typical situations in which optimization algorithms are ineffective are presented.The reasons that may cause algorithms to be ineffective are analyzed.(4)The efficacy of several commonly used optimization algorithms when solving different BEO problems is evaluated using the efficacy evaluation indices proposed in the paper.Besides,suggestions for choosing proper optimization algorithms are provided according to the properties involved in BEO problems.(5)A new method for determining the best parameter settings of algorithms for specific BEO problems is developed.By using this method,the best parameter settings of three suggested optimization algorithms are determined for different BEO problems.In addition,suggestions for reasonably setting algorithm parameters in BEO are provided.(6)The BEO technique is applied to the optimization design of a Tourist Service Center at the east entrance of Niushou Mountain Scenic Spot in Nanjing.The recommended optimization algorithm and the ones that are not recommended are used in the case study to compare their efficacy.Results indicate that the efficacy of the recommended optimization algorithm is significantly better than the others and can better solve BEO problems in actual design practice.