Precise Relative Positioning for Tandem Drifting Cars

Drifting provides insights into how autonomous cars handle excursions past stable limits. So far, autonomous drifting has been demonstrated for a single car where GPS carrier phase measurements provide cm-level accuracy in absolute position via base station corrections. We aim to extend the single car drifting to tandem drifting. However, in autonomous tandem drifting, it is hard to guarantee cm-level accuracy in relative position using standalone carrier phase measurements since noisier drift dynamics create a large search space of integer ambiguities which makes integer fixing challenging. We propose a precise relative positioning framework for tandem drifting where we use drift dynamics to aid carrier phase measurements. Utilizing drift dynamics, we create an a priori estimate of relative position that constrains the physical search space, and the induced integer search space. To improve integer ambiguity fixing rate, we adaptively adjust the integer search space based on the noise in drift dynamics. We demonstrate our algorithm’s performance on a simulated tandem drifting case. Our algorithm achieves cm-level accuracy even with noisy GPS measurements and shows improved positioning over conventional methods such as LAMBDA.