Model And Method For Ratbot Automatic Control

Bio-robot is a new type of robot systems, which is built based on Brain-Computer Interfaces (BCI). Micro-electrical stimulations are delivered to the animals’ motor and sensory cortex through a set of microelectrodes to make them have ’virtual feelings’. Then the animals will do the behavior follow the ’virtual feelings’.Nowadays, the research about bio-robots is focused on manual control. Due to the lack of its automatic control, bio-robots can’t be used in practical applications. Bio-robots have self-consciousness, so the control methods of traditional robot are not suitable for them.We use ratbot as research object, considering the movement intent of the rat. From the aspect of path planning, calculation of motion states and automatic navigation control logic, an automatic navigation system is achieved. The content of this paper includes the following aspects:(1) Using image processing methods to build the topology map of the eight-arm experimental environment automatically. The rat’s self-path selection is seen as priori probability, which will combine with probabilistic graphical models to calculate the ratbot’s route.(2)The motion is a continuous concept. Sparse optical flow, based on successive frame, is used to extract motion states of ratbot. Compared to the single frame processing method, the result can reflect the motion trend of ratbot more accurately.(3)Because of the neurophysiological characteristics of ratbot control, control instructions transmitted frequency is restrained. Furthermore, if the ratbot turn successfully, the rewarding stimulation must be transmitted in time to strengthen operant conditioning. So we split the manual navigation states, considering the time constraints, build the control logic model of automatic navigation through timed automata. The UPPAAL tool is used to do related validation.Finally, two different experimental paradigms are designed. Ratbot can be automatic controlled in the eight-arm maze environment. The results show that the automatic navigation system which considers the movement intent of rat has good performance.