TU-TESL

• 22

In dynamic computer games, computers are also decision makers. Players may often find themselves competing with games to achieve their goals. Similarly, when driving a motor vehicle equipped with technology, a human driver sometimes scrambles with the car for the steering wheel to ensure the safety of himself and others and eventually reach his destination. So far, most of the research on this kind of driving interaction is based on the theory of uncooperative game, that is, the decision-making of drivers and computers on how to drive a car is not consistent. (photo source: cc0 public domain) according to foreign media reports, contrary to many previous studies, in order to better understand and predict the outcome of the dilemma of steering wheel control, Dr. Xiaoxiang Na and Dr. David J. Cole from Cambridge University think that it may be more appropriate to use cooperative game theory to model this kind of interaction problem. Using cooperative game theory, they construct scenarios based on Pareto equilibrium theory (that is, it is impossible to improve the interests of one party without harming the interests of others). In this equilibrium state, neither human drivers nor auto steering systems are willing to unilaterally change their steering behavior. In Pareto state, although the driver and the automatic steering system have different ideas about how to run the car, for example, in order to avoid hitting pedestrians, whether to go straight or change lanes, the driver will agree with the intention of the automatic steering system and make a decision accordingly. In order to prove whether cooperation can improve the experience of autonomous driving, the researchers also compared the modeling results with the results of non cooperative game. They pointed out that the adoption of the cooperative strategy "strengthens the driver's steering angle control, so that the car can avoid dangerous lane changes and safely and quickly go straight." In the future, researchers need to conduct more experiments to further study the impact of cooperative strategies on the interaction between human drivers and automatic steering system. They hope that the strategy can be implanted into the automatic driving technology, so that the automatic driving technology takes into account the real driving experience of the human driver, thus optimizing the sharing control between the human and the automatic driving vehicle.