The invention discloses a method for constructing an interpretable machine learning-assisted channel model in mixed traffic, and the method comprises the steps: constructing a mixed traffic scene, and setting the input features and output features of the channel model; the method comprises the following steps: constructing a mixed traffic channel data set T, dividing the mixed traffic channel data set T into two sub-data sets T = {TV, TI}, and respectively dividing the two sub-data sets into a training set and a test set; the method comprises the following steps: respectively training a training set of a vehicle-to-vehicle V2V communication data set TV and a training set of a vehicle-to-infrastructure V2I communication data set TI by using a machine learning ML algorithm LightGBM, and constructing a vehicle-to-vehicle V2V channel model and a vehicle-to-infrastructure V2I channel model in a mixed traffic environment; the method comprises the following steps: respectively analyzing a V2V channel model and a V2I channel model by using a test set of TV and TI by adopting a Shapley additive explanation SHAP method, acquiring key features which have obvious influence on communication quality according to an SHAP value, generating interpretable models g (xV ') and g (xI'), and constructing a lightweight mixed traffic channel model; according to the invention, the safety and high efficiency of traffic transportation are realized, and powerful support is provided for hybrid traffic communication.
一种混合交通中可解释性机器学习辅助的信道模型的构建方法,构建混合交通场景,设置信道模型的输入特征和输出特征;构建混合交通信道数据集T,分为两个子数据集T={TV,TI},并分别划分训练集和测试集;使用机器学习ML算法LightGBM分别对车对车V2V通信数据集TV和车辆到基础设施V2I通信数据集TI的训练集进行训练,构建混合交通环境下的车对车V2V信道模型和车辆到基础设施V2I信道模型;采用沙普利加性解释SHAP方法,利用TV和TI的测试集分别对V2V信道模型和V2I信道模型进行分析,根据SHAP值,获取对通信质量产生显著影响的关键特征,生成可解释性的模型g(xV′)和g(xI′),构建轻量级的混合交通信道模型;本发明实现了交通运输的安全性和高效性,为混合交通通信提供了有力的支持。
Construction method of interpretable machine learning assisted channel model in mixed traffic
一种混合交通中可解释性机器学习辅助的信道模型的构建方法
2024-02-13
Patent
Elektronische Ressource
Chinesisch
Interpretable traffic flow prediction stacking method
Europäisches Patentamt | 2024
|Interpretable machine learning approach in estimating traffic volume on low-volume roadways
DOAJ | 2020
|Man-machine mixed driving traffic flow vehicle lane changing model construction method
Europäisches Patentamt | 2022
|Interpretable space-time analysis method for traffic jam prediction
Europäisches Patentamt | 2022
|