Abstract. The robot impedance control performance decreases with unknown or changing environmental stiffness and damping parameters, in order to resolve this problem, this paper designs a self-adaptive robot impedance control method, which is characterized by integration of off-line learning and on-line adjustment to afford the stiffness and damping of the robot control system's impedance model competent for unknown or changing environment. For the off-line learning, defining the robot impedance control performance criterion, and establishing the geometric representation of the varying stiffness parameter, we derive the initial values of stiffness for the impedance model. Further, a neural network is designed to estimate the environmental effective stiffness, and combined with critical damping condition of the second-order robot-environment interaction system, we solves the initial value of damping for the impedance model. During the on-line adjustment, a rule self-tuned fuzzy controller is dedicated to adjust the stiffness and damping of the impedance model based on robot real-time contact force and position feedback. At last, experiments demonstrate the excellent stability and accuracy for the robot-environment contact force tracking control.

No abstract

Controller is the core of robot control system. This design uses a dual stc89c52rc single chip microcomputer control, which processes and responds very quickly,with the campaign, obstacle avoidance, automatic tracing function. This article describes the hardware and software design of control system.

No abstract

Location technology as Wireless Sensor Network’s support technology for most applications has been widely researched, ZigBee is one of representative technology of WSN. On the basis of the common two-step localization algorithm, this paper proposes a locating method that optimize the measurement data using genetic algorithm before locating calculation, then using multilateral measurement least-square method for calculating position. Experimental results show that the modified algorithm is significantly reduced the position error, effectively improve the position accuracy.