基于模糊着色Petri网的车辆运动轨迹预测

车辆运动轨迹的准确预测可以对潜在交通冲突进行实时有效预测，并为解决交通冲突提供最佳策略。以真车实验数据为基础，分析车辆运动轨迹的影响因素，建立车辆运动轨迹方程，根据坐标变换以及各个方程变量参数之间的转换关系，将车速纵向和横向加速度、车身侧倾角以及俯仰角速度作为模型输入，构建论域以及隶属度函数。此外，根据车辆不可能在较短时间内状态发生急剧变化建立相应的模糊规则，建立模糊着色Petri网模型，对车辆运动轨迹进行预测。以真实路车实验数据对模型进行训练与测试，测试结果表明该Petri网模型能够对短时长内的车辆运动轨迹进行有效预测。

Prediction of vehicle trajectory based on fuzzy colored Petri net

The accurate prediction of vehicle trajectory can predicate the potential traffic conflict in real time and provide the best strategy to solve the traffic conflicts. The influence factors of vehicle trajectory are analyzed and the vehicle trajectory equations are established. According to the coordinate transformation and the relationship between variable parameters of each equation, the longitudinal and lateral acceleration are taken as model inputs and the domain of discourse and the membership functions are determined. In addition, the corresponding fuzzy rules are established on the basis of the fact that the states of vehicle may not change in a relatively short time. The colored fuzzy Petri net model is established to predict the vehicle trajectory. The real experimental data are used to train and test the model and the results show that the Petri net model can effectively predict vehicle trajectory within a short time.