Study On The Residencegenerative Design Forhunan Based On Environment Characteristics In Hilly Area

As a new medium to integrate architecture with environment, space, function, construction and art, generative design (GD) shows a satisfactory adaptation ability and excellent development potential in dealing with complex problems in urban design, which enriches the approach to architecture design. It is commonly used in the field of architecture, which has become a research focus. The critical issue faced at present is to figure out the application method, application range and application achievement.Architectural generative design includes many generative theories and methods, which require to select and integrate when designing. Based on the theory and methods of generation, Hunan residence which featured by hilly region environment was chosen as the case study, which aims to study how to put these theories and methods into practice under different design conditions and requirements, thereby new design methods and strategies can be developed. In the present study, mathematical modeling was utilized to construct the generative design process, which makes the entire generative process possible. Different design methods were coordinated to the changed requirements. These design methods cannot only be applied to the hilly residential area of Hunan province, but to some extent, can also be applied to the further research objects. It has significant implications for theory and practice. The present study mainly includes the following contents:Firstly, environmental characteristics such as solar radiation, wind field, sunshine duration and terrain of Hunan hilly region were analyzed. The process of the formation, evolution and development of architectural generative design were overviewed. Besides, some typical generative design methods such as Genetic algorithm (GA) design, Self-organizing generate design, and Complex adaptive generative design and Intelligence generate design were illustrated to provide a research foundation for generative design for the hilly area of Hunan Province.Secondly, the distribution of solar radiation in the hilly areas on the buildings was analyzed. Moreover, according to the GA, the residential prototype was extracted, transforming readable input data and function model for the computer recognition. The shape coefficient of building and dwelling size optimization function was chosen as the fitness function. Therefore,135 generative results were obtained when the constraint condition was converted to generative object in the pre-set generation rules. Preliminarily, six kinds of graphic design programs were selected, particularly, one program was further optimized, and whose shape coefficient was calculated to be around 0.30. The greatest strength of the generative design program is that it can generate a large number of schemes in a short time, and control the shape coefficient within a set range, which contributes a lot in reducing residential energy consumption.Thirdly, five prototypes of collection of residential units were selected. The corresponding generative rules were set under the precondition that every collection residential units could share and maximize the utilization of the courtyard. By using computer models, the optimal combination of the courtyard and residence was designed, and also the suitable direction of the courtyard for different parts of the hill was determined. Then, the generative model was optimized. The residential and courtyard model generation could not only improve the living environment of the high-rise residential building, but also will make multi-family buildings share the same courtyard. The aggregate building area of the courtyard was optimized by a computer, besides, the proportion of the total building area was in control while maximizing space utilization.Fourthly, take Changsha for example, the characteristics of the sunshine environment in Hunan hilly region from 1987 to 2013 were analyzed, the trend of the annual change was achieved. The relationship among the distance for sunlight, the sunshine duration, and the orientation of residence was further analyzed. The corresponding complex sunshine adaptive model was developed according to the amount of Changsha sunshine. The sunshine model can be set up according to the analysis of the solar altitude and solar azimuth with the aid of computer technology. Each system unit was coded into the graphics that the computer can recognize, operational, and output. Seven generative rules were generated. Then, the residential planning model was generated, which can be fully consistent with the requirements of land specifications, when the sunshine is maximized.Finally, the topography feature in Hunan hilly region was analyzed. The bidirectional relationship between the residence and the topographic factor was discussed. What’s more, the terrain adaptability arrangement for residence was studied. The generative design of an intelligent system which is adapted to the terrain of Hunan hilly region was applied to optimize the residence layout planning and design.