Method for realizing movable environment based on unity ML plug-in

Abstract

The invention discloses a method for realizing a movable environment based on a unity ML plug-in, and the method comprises the steps: 1, building a movable training scene of a vehicle test through employing a Unity 3D development assembly, importing a dynamic object in the training scene through employing a Vehile Tools resource package, and setting a motion state for the dynamic object; 2, creating an intelligent agent in the movable training scene to serve as a target training vehicle; creating a virtual sensor for the target training vehicle; 3, constructing a training model, and training the target training vehicle by using the movable training scene to obtain a trained model file; and 4, importing the trained model file into unity engineering, reading the model file by TensorFlowSharp, and applying the model file to a Brain vehicle object, thereby realizing automatic movement in a specific scene. According to the method, reinforcement learning or other machine learning is carried out based on an AI framework tensorflow and an ML plug-in, the surrounding movable environment is simulated, and a certain autonomous intelligent environment object is obtained through training.

Description

technical field [0001] The invention belongs to the technical field of automobiles, and in particular relates to a method for realizing a mobile environment based on a unity ML plug-in. Background technique [0002] At present, in some project activities that require 3D display, such as urban planning, real estate virtual real estate, etc., it is often necessary to realize the movement of some surrounding road vehicles to improve the realism and sense of substitution of the display. The traditional method is to set the moving path of surrounding objects in the scene, and then randomly select the path through configuration, randomly generate different vehicles, and randomly set the speed direction to achieve a more realistic surrounding environment. However, the randomness of this method is too small, and it is often easy to see that the object moves along certain specific trajectories; at the same time, when encountering congestion and other situations, the object still move...

AI Technical Summary

Problems solved by technology

However, the randomness of this method is too small, and it is often easy to see that the object moves along certain specific trajectories; at the same time, when encountering congestion and other situations, the object still moves along the pre-set trajectory, which cannot naturally simulate special situations. lead to poor authenticity

Images