Method for quickly recommending similar house types based on house type contours and distribution characteristics

Abstract

The invention provides a method for quickly recommending similar house types based on house type contours and distribution characteristics, and relates to the technical field of house type data processing. The method comprises the following steps: S10, preparing a data set for training and matching recommended house types; S20, designing a data model, through the designed data model of a unified building information model, storing structure and space information of a house type design; S30, extracting house type features, designing the features of the house type, selecting the features of the contour graph, the area, the house type distribution, the space distribution and the space number of the house type, and extracting the features of the data set; and S40, outputting a mathematical model, training the algorithm mathematical model by using the training data set and a neural network method to obtain a mode estimation parameter, and fitting a matrix to obtain the mathematical model. The method has the beneficial effect that an accurate reference design scheme is provided for a house type designer.

Description

technical field [0001] The present invention relates to the technical field of house type data processing, and more specifically, the present invention relates to a method for quickly recommending similar house types based on house type outlines and distribution features. Background technique [0002] At present, when designers in the industry design house types, they need to refer to house types with similar outlines and similar distributions. Through experience or memory, they manually search in the design archives of their design institutes, teams or individuals to find similar design solutions. The scheme library may be the entity of the design drawing, or it may be an electronic drawing that has been digitally archived, but all of them require manual retrieval. Manual retrieval has the following defects: 1. Rely on the designer's design experience and design experience; 2. It takes a lot of time 3. Limited by the amount of data in the design scheme database, the search ...

AI Technical Summary

Problems solved by technology

The scheme library may be the entity of the design drawing, or it may be an electronic drawing that has been digitally archived, but all of them require manual retrieval. Manual retrieval has the following defects: 1. Rely on the designer's design experience and design experience; 2. It takes a lot of time 3. Limited by the amount of data in the design scheme database, the search result set is limited; 4. Limited by manual judgment, the results may not be accurate enough

[0003] Algorithmic matching technology is also limited by the inconsistency and insufficient quantity of data sources, and there is no unified building information model, resulting in unsatisfactory accuracy and regression rate of the results

Its main defects are manifested in the following aspects: 1. There is no perfect and unified building information model, which is used to store all kinds of building and house type data in a unified manner, and provide perfect information at the same time; 2. It cannot support data sources in various formats. Such as automatic recognition of CAD (drawing software) drawings, etc.; 3. The matching of the graphics of the top view of the house type only considers the characteristics of a single space, and does not consider the characteristics of the relative relationship between spaces, as well as the characteristics of space types and areas.

Images