Reinforcement Learning and Time Perception -- a Model of Animal Experiments