Urban three-dimensional wind speed prediction method based on three-dimensional morphological characteristics and machine learning

By constructing a three-dimensional wind speed prediction method for cities based on three-dimensional morphological features and machine learning, the problem of existing technologies being unable to accurately characterize the differential impact of building spatial morphology at different heights on airflow is solved. This method achieves efficient three-dimensional wind speed prediction across the entire height range of the urban canopy, improving prediction accuracy and cross-scenario adaptability.

CN122154498AActive Publication Date: 2026-06-05SHANDONG UNIV OF TECH
5 Cites 0 Cited by

Patent Information

Application Number
CN202610628668.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-05-09
Publication Date
2026-06-05
Estimated Expiration
2046-05-09

AI Technical Summary

Technical Problem

Existing urban wind speed prediction methods are unable to accurately characterize the differentiated impact of building spatial morphology at different heights on local airflow, resulting in the inability to achieve three-dimensional wind speed prediction across the entire height range of the urban canopy, thus limiting the model's cross-scenario adaptability and engineering application value.

Method used

By employing a method based on three-dimensional morphological features and machine learning, a three-dimensional wind speed level prediction model for cities is constructed by building a vertically hierarchical computing system and regular grid prediction units, combined with a random forest classification algorithm. This model accurately depicts the differentiated impact of building morphology on airflow and achieves efficient prediction of the three-dimensional wind speed field.

Benefits of technology

It improves the accuracy and computational efficiency of wind speed prediction, reduces computational costs, has excellent cross-scenario adaptability, and can provide efficient and reliable three-dimensional wind environment analysis support for urban planning.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention belongs to the technical field of wind speed prediction, and particularly relates to an urban three-dimensional wind speed prediction method based on three-dimensional morphological characteristics and machine learning, and the method comprises the steps: obtaining an urban three-dimensional building data set of a research area; dividing the research area into a plurality of regular grid prediction units, and establishing a spatial mapping relation; forming a three-dimensional prediction unit set jointly determined by the regular grid prediction unit and the prediction height layer; constructing an original three-dimensional city form feature set; performing direction unification processing on the original three-dimensional urban form feature set to obtain a direction unification model input sample, and constructing an extended wind direction sample; and adopting a random forest classification algorithm to construct and train a wind speed grade prediction model, obtaining a final urban three-dimensional wind speed grade prediction model, taking a direction unification model input sample of an actual research area as input, and outputting a wind speed grade prediction result. The method can achieve the efficient prediction of the three-dimensional wind speed field of the urban canopy, is high in precision, is quick in calculation, and is high in cross-scene adaptability.
Need to check novelty before this filing date? Find Prior Art