Research On Robot Shared Control Based On Trust Model

In the process of development and utilization of natural resources,teleoperation robots are playing a more and more importance role.Because of the limitation of control in robotics,it is difficult to design a completely autonomous robot in the extreme environment immediately.Hence,the teleoperation robot is becoming increasingly important in this case.Combing the automatic control and manual control is the current research trend.Shared control falls between supervisory control and bilateral control,the human operator can intervene in an autonomous task executed by the robot and the robot can augment the direct commands generated by the operator.Human has the ability to control,and receive feedback from the remote robot.In the field of robot shared control,there are two kinds of typical shared control mode,one is that operator and robot control the different degree of freedom respectively,the other is that the operator manipulator the robot to the target location in advance,then the robot completes the task automatically.The degree of fusion between human intelligence and robot intelligent are still low and relatively independent.In order to solve the fusion problem in shared control,the article first has built visual virtual guiding force based on artificial potential field,then,through the fuzzy logic system based on human-robot trust model,the weight coefficient of operator and robot can be got.Finally,the operating force and virtual guide force multiplied by the corresponding weight coefficient respectively,fusion force can be got by adding them,haptic device and robot is controlled by the position movement trend predictor synchronously.This paper has mainly completed the following works:1.Calculating the kinematics of the haptic device and the robot,designing the position and pose mapping relationship between haptic device and robot coordinate.2.Human robot trust model has been built.The model shows how operator,robot and environment influence human-robot trust relationship in detail.The teleoperation system must guarantee that human-robot trust relationship should keep in an expected range.3.Based on fuzzy logical method,the index of human-robot trust relationship and the condition of robot as input,the weight coefficient of operator and robot can be got through fuzzy inference system.4.The shared control strategy based on data fusion has been built.The fusional force is used to control haptic device and robot.5.Writing the control program in VS+Qt environment with the method of multithreaded programming.Sub-function modules are scheduled by a proper time sequence.The effectiveness of the control method was verified by the experiments.