If you are trying to locate by using TDOA (time difference of arrival), would the sensors not need to be a few feet apart to get a measurable time difference? How far are the sensors apart?
What about sensing peak intensity or how 'loud' the chirp is to derive a general directional bearing? So you would poll their last reading and adjust heading so the front sensor has the highest reading? You could then use an ultrasonic rangefinder to provide a distance measurement and collision avoidance.