【源码】驾驶员辅助驾驶场景中基于Agent的建模演示

基于Agent的建模（ABM）是一种通过模拟给定场景中大量自治Agent的行为和交互来研究和预测复杂群体行为的建模方法。

Agent-Based Modeling (ABM) is a style of modelling to investigate and predict the emergence of complex group behaviors through simulating the actions and interactions of a large among of autonomous agents in given scenarios.

ABM在学术界得到了广泛的应用，如博弈论、复杂系统和计算社会学等。

ABM has been extensively used in the academia, e.g., game theory, complex system and computational sociology.

对于自动驾驶，ABM也是一种流行的模拟方法，例如，在制定驾驶政策和安全验证时。

With respect to autonomous driving, ABM is also a popular simulation approach, e.g., in developing driving policies and safety verification.

传统的运动规划和控制方法，如PID控制、反馈线性化或模型预测控制，期望对其他交通参与者的未来轨迹进行预测，以避免碰撞，但实际交通场景涉及各种道路用户之间的复杂交互（Schwarting, Planning and Decision-Making for Autonomous Vehicles, MIT, 2018）。

The conventional motion planning and control methods, e.g., PID control, feedback linearization or model predictive control, expect a prediction over the future trajectories of other traffic participants in order to avoid collisions, but real traffic scenarios involve complex interactions among various road users (Schwarting, Planning and Decision-Making for Autonomous Vehicles, MIT, 2018).

为了克服这一挑战，新兴趋势是行为感知的运动规划和基于学习的方法。

To conquer this challenge, the emerging trends are the behavior-aware motion planning and learning-based approaches.

ABM可以很自然地用来处理复杂、杂乱的环境，同时对彼此之间的不确定交互进行建模。

ABM is then could be naturally exploited to handle the complex, cluttered environments while modeling the uncertain interactions with each other.

这将使Simulink成为开发和验证基于学习的算法的集成平台。

This would expend Simulink as an integrated platform to develop and verify learning-based algorithms.

有关说明，请参见以下视频：

For a description, you can see this video: https://www.youtube.com/watch?v=nKOleGGEwJI

更多精彩文章请关注公众号：