Each individual is characterized by a two-component Rayleigh distribution, which reflects two modes of movement: active movement (progression) and sedentary state (drift). A mechanical analogue would be the motion of a Brownian particle with a variable mass: small during active movement and large during drift (+ "decision-making load"). A stochastic model with random switching between these modes is constructed, which makes it possible to estimate the amount of "added mass". The experimentally obtained autocorrelation function of velocity also demonstrates two-phase attenuation with two characteristic times corresponding to the duration of progressions and the characteristic time of switching between modes.
Averaging individual distributions over a group yields a simple exponential distribution due to individual differences in the characteristic velocities. A mechanistic analogue of the movement of an individual rodent would be movement with viscous friction (deceleration is proportional to speed), whereas the group-averaged model behaves as if it were affected by dry friction (constant braking). The proposed stochastic model and algorithm for creating an averaged bimodal "reference" distribution can be useful for extracting information about locomotor modes, for example, in biomedical research. Characteristic speeds can be a sensitive parameter for assessing the effects of psychoemotional factors, side effects of drugs, neurological problems with movement initiation, etc.