Optimization Of Building Design Parameters For Energy Efficient Of Naturally Ventilated High-rise Residential Buildings In Tianjin

Natural ventilation can not only improve indoor air quality,but also reduce the energy consumption of residential buildings in summer.However,it's very difficult to apply the optimizations that integrate natural ventilation and energy saving in the architectural design,due to the complexity of natural ventilation and some constraints between summer cooling and winter heating.So,it was very necessary to study the quantitative relationship between the architectural design parameters and the energysaving performance of natural ventilation which can benefit the application of these passive techniques in architectural design and enhance the residential building energy performance.This paper analyzes the design characteristics of residential buildings in cold climate and the impacts for natural ventilation.The statistics of new residential types that were built in recent seven years(2010 ~ 2016)were conducted using ArcGIS and result shows that slab-type high-rise residential buildings,which floor numbers between 12 and 33,are accounted for the vast majority among all the residential building types,and the internal function of this type buildings were counted.Based on these statistics,the design parameters calculation model of slab high-rise residential coupling heat and air flow models was established by using software such as Rhino,Grasshopper,Archsim and Energyplus.Using the calculation model,energy simulations were carried out according to the six key design parameters(standard floor area(A),aspect ratio(R),building floor numbers(L),north window to wall ratio(N),south window to wall ratio(S),building orientation(D),with and without natural ventilation.The influence of natural ventilation on the building loads was analyzed.It was found that the ventilation rate had great influence on the building load,and the "dynamic air change rate" should be used in building energy consumption simulation,instead of the “constant one”.In order to reduce the optimization time and overcome the shortcomings of the joint simulation platform,the surrogate model of building loads considering natural ventilation(BLSNVM)was established with the MATLAB BP artificial neural network,which was trained by using 2001 sets of random data samples simulated with the calculation model.The prediction accuracy of the model was analyzed.The data showed that the tolerance scope is less than 12%,which satisfies the evaluation of residential energy consumption in the design stage.Finally,optimization model(NNGA)of the design parameters was established by integrating the BLFNVM and genetic algorithm.An optimal solution parameter system(OSPS)according to A and L was established using NNGA.The relevant architectural design parameters can be easily optimized through NNGA and OSPS,which were verified with two practical cases.