"The system works by simulating human driving logic," he wrote in a simple reminder. "Then, by optimizing the priority criteria and improving the yielding mode, we enable the group of people who need to drive fast to reach their destination in the shortest time, and the effects on the rest of the people are recursively applied. It is equivalent to assuming that every car has, uh... noble morals, and improving the efficiency of the entire system through yielding layer by layer."
"But this may not be the optimal solution. What I mean is that it can achieve the highest efficiency under normal human logic, but it is not the ceiling." He paused: "It is just a hypothesis. We incorporate road rules, traffic light flashing time, and road speed limits...all these things into the autonomous driving system, and add them as variables to the system scheduling. For example, speed limits can be opened in advance during peak hours, or traffic light times can be dynamically adjusted according to the number of waiting vehicles. I don't know much about traffic theory, so my ideas are very limited. But the central idea is that when a car starts this autonomous driving trip, all traffic rules will give way to it to a certain extent."
After a brief silence, Jiang Yun said, "This problem has been considered initially, but it is not mathematically feasible."
He said: "When local rules are changed, this change will cause disturbances in other parts of the system... You should understand. In this case, if you want to perform overall calculations, the level of calculation is far beyond what we can achieve now."
Another person beside him wrote a line of formula on a white paper, with a hint of reproach in his voice: "Ripple theory, when there are multiple ripples on the water surface, they will eventually cancel each other out or become a mess. In this case, the calculation will fall into chaos, and the final efficiency will be even lower than manual logic."
Login to comment
Be the first one to comment...