Section: New Results

Using Fractional Calculus for Cooperative Car Following Control

Participants : Carlos Flores, Vicente Milanés, Fawzi Nashashibi.

In the field of Advanced Driver Assistance Systems (ADAS), there are two main types of systems: passive and active ones. Specifically the active ADAS, they are capable of taking partial or complete control of the vehicle. Among these techniques, Car-Following has arisen as one important solution to traffic jams, driver comfort and safety.

Scoping on the evolution of the control involved in Car Following, it can be remarked the improved version of the cruise control system, Adaptive Cruise Control (ACC). This system allows the vehicle to maintain a desired distance gap measured by raging sensors (LiDAR, radars, etc), by controlling longitudinally the vehicle through the throttle and brake.

Afterward, the addition of Vehicle to Vehicle (V2V) communication links allowed the vehicles to maintain even shorter distances between each of the string members, by performing a Cooperative ACC (CACC). Focusing on CACC formations, a control structure must be conceived to guarantee stability and string stability as well. As a core of the control structure, the controller must be able to maintain the vehicle in the desired spacing in a stable, robust and comfortable.

Towards achieving this goals, it is proposed to use fractional order calculus to gain a more flexible frequency response and at the same time satisfy more demanding design requirements. This mathematical has been used for years for different applications providing good results and outperforming classical techniques in the industrial control field, due to its capability of describing systems more accurately than integer order calculus. Several research lines are stated to achieve these objectives:

• An exhaustive identification process of the experimental platforms dynamics. Allowing further comparison between the empirical identified dynamics of the real vehicle and a theoretical mathematical dynamic model. Such permits to design much more effective and stable control algorithms for both the lateral and longitudinal command of the vehicle.

• Conception of a Car-Following gap regulation controller using fractional order calculus, which has been proven that yields a more accurate description of real processes. The controller should satisfy more demanding design requirements [31], allowing to extend the scope of Car Following controllers' design. This controller should be framed into an appropriate control structure both for ACC and CACC

• Further investigation on the effects of communication delays and latency in the V2V links, as well as study different control structures that react not with the preceding vehicle's behavior but also other string members.