A synchronous control strategy of robot social behavior driven by scenario information and neural modulation mechanism.

Abstract

Robots have been widely employed in scenarios that involve various environmental factors and social individuals. As one kind of social companion, robot is supposed to obey human social protocol and display anthropomorphic behaviors. In this paper, we focus on the problem of robot behavior control in multi-individuals scenarios, and build a coordinated robot behavior model containing body movement/orientation, head rotation and eyeball movement. Within the proposed model, a synchronous control strategy based on social space theory and neural modulation mechanism is proposed. This strategy collects RGB-D camera stream and acoustic field data perceived from multi-individuals scenario, and controls the robot to complete movement and social gaze behaviors. As for the eye-head coordinated gaze behavior, it is modulated by a novel optimal control algorithm based on the minimum neural transmission noise. Above works are validated on the Xiaopang robot platform, the experimental observations indicate that the robot can achieve anthropomorphic response in dynamic multi-individuals scenario. Within above promising results, the effectiveness of this strategies could be proven.