4 results about "Street network" patented technology

The application discloses a flood simulation method based on urban infrastructure and a device thereof, and the method comprises the following steps: collecting street view images of a city, wherein the street view images comprise street view data collected from different urban infrastructures at different angles and different times; classifying and processing the street view data according to the types of the infrastructures, and extracting urban infrastructure data; fusing the infrastructure data and terrain data of the city to obtain fused terrain data of the city; dividing catchment areas of the city based on the fused terrain data and extracting water system data, and establishing a flood simulation model based on urban infrastructure according to the water system data, so that the influence of urban street networks, fence systems, buildings and other infrastructures on urban flood processes is fully considered, the simulation accuracy of a city rainstorm flood model is improved, and urban flood disasters are reduced.

The application provides a historical urban area potential street vector identification method based on axis extraction. The method comprises the following steps: step one, historical urban area map data acquisition and processing; step two, division of calculation units and preliminary evaluation; step three, street axis extraction based on a triangular mesh and parameter rules; and step four, street axis optimization and output. The application can set rules for the existing planning framework and street layout, explore potential street axis lines with traffic conditions in the historical urban area, and screen the generated results according to the parameter rules, thereby guiding the generated results. The application can provide quantitative data support for designers to optimize the street network structure without destroying the existing texture of the historical urban area.

This invention discloses a collaborative improvement method for horizontal wind speed algorithms in urban street network models. The method includes using high-resolution CFD wind field simulations as near-true values, comparing and verifying wind speeds at different locations in the urban street network model with the near-true wind fields, introducing local roughness and displacement height parameters incorporating building characteristics, and systematically evaluating the impact of combinations of local roughness and displacement height, along with CFD-derived local friction velocity, on wind speed simulation at different scales. Based on the evaluation results, the optimal parameter combination is selected to replace the regional average roughness and displacement height parameters, optimizing the wind speed simulation capability of the urban street network model under non-uniform urban underlying surface conditions. The advantages are: by combining the CFD-derived local friction velocity with local roughness and displacement height parameters, the non-uniformity of the underlying surface height in the city center can be more accurately characterized, thereby precisely simulating the drag effect of urban buildings and helping to obtain horizontal wind speeds closer to reality.

The invention discloses a multi-type street network walkability dynamic optimization method based on reinforcement learning, and the method comprises the following steps: obtaining multi-source road network data, and recognizing and classifying different forms of street networks; setting a two-level action space of general structure optimization and type exclusive optimization according to the network characteristics of various types of streets; a two-dimensional walkability evaluation index system fusing the dynamic behavior result and the static road network structure is constructed, and index fusion and standardized evaluation are achieved; an autonomous iterative optimization mechanism is constructed based on reinforcement learning, a walking comprehensive score is used as a reward function, and an agent autonomously decides and dynamically adjusts a road network state; and realizing three-dimensional visual real-time rendering and multi-terminal interactive comparison of the optimization scheme, and outputting a Pareto optimal strategy set. The problems that in a traditional method, the evaluation dimension is single, optimization depends on manpower, simulation and optimization are disjointed and the like are solved, and efficient, accurate and systematic improvement of the multi-type street network walkability is achieved.