Ding Liu

Real-time recognition of road traffic sign in motion image based on genetic algorithm

In this paper, a new technology based on the genetic algorithm and new image filter for recognizing road traffic sign from motion image captured by a CCD camera in a car was developed. In order to realize a real-time position recognition, the step genetic algorithm with search region limits and image filter (denoted by SVF) were proposed. The genetic algorithm with search region limits was employed to detect the position and size of traffic signs in real-time as well as SVF was employed to extract specified colors. The feasibility and validity of the proposed scheme are demonstrated through road driving experiments.

Robot trajectory tracking control based on fuzzy immune PD-type controller

Based on the dynamic nonlinearities of robot manipulators, a fuzzy immune PD-type control algorithm is proposed for robot trajectory tracking. The algorithm combines fuzzy control, immune feedback mechanism of organism with conventional PID control. In the algorithm, the suppression number of suppression cells is described by a nonlinear function, which is approximated by a 2D fuzzy controller. The given algorithm can overcome the influence of modeling error and parameter varying when robot dynamic model parameters change from 10% to 190% randomly. Simulation and experimental results of 2DOF robot manipulator show that the control scheme has better tracking precision, stronger robustness and superior control performance to conventional PD controller.

GA-based object recognition in a complex noisy environment

This paper describes a method for object recognition in a complex noisy environment based on the genetic algorithm (GA). A small object is represented by their binary edges. A fitness function is constructed by the shape of an object in combination with its frame model to search for the position and orientation of the target in the input image. In order to enhance the orientation function of the fitness function, some preprocessing operations have been done. The simulation result shows that the method presented is effective and has great practical value.

Robot end-effector 2D visual positioning using neural networks

A visual positioning controller of robot manipulator system with a camera in hand is presented in this paper, where a feedforward neural network is involved to drive the end-effector of manipulator to the desired position instead of the proportional controller. In this case, the visual sensory input is directly translated to world actuator domain. Simulation results show that this method can drive the static positioning error to zero quickly and keep good dynamic response at the same time compared with proportional control law.

Robot manipulator controller based on fuzzy neural and CMAC network

A new controller is proposed to deal with the uncertainty in robot manipulator dynamic system. This controller is composed of fuzzy neural network (FNN) controller which replaces computed torque controller, cerebellar model articulation (CMAC) controller which compensated control error online and slide mode controller which compensates fuzzy neural network fit error to enhance robustness of the control system. It is shown that this controller outperforms conventional fuzzy controller by improving learning ability that the conventional fuzzy system does not have with simulations of a two-link manipulator.

Path planning algorithms for power transmission line inspection using unmanned aerial vehicles

The effectiveness and efficiency of using unmanned aerial vehicle (UAV) for automated power transmission line inspection is tightly related to the inspection paths designed for UAV. Different types of UAV are suitable for different inspection tasks and have different requirements for path planning. Based on the contents of the transmission line inspection, it can be divided into tower monitoring and line corridor monitoring. First, according to the characteristics of the tower monitoring, the multi-rotor UAV is used for the tower inspection. By considering the safe distance bewteen UAV and the tower and the features of the camera, the genetic algorithm (GA) is used to design a rational inspection path. Then, according to the requirements of the line corridor monitoring mission, the fixed-wing UAV is used for long distance inspection and the path planning mathematical model and objective function are established. Polar coordinate coding is used to overcome the restrictions of the maximum path deflection angle, the minimum step length and the number of the maximum path nodes. The GA and genetic simulated annealing (GSA) algorithms are used to obtain effective inspection paths. Simulation results show that the proposed optimization scheme in this paper can find the optimal inspection paths.

Nonlinear tracking control methods applied to Qball-X4 quadrotor UAV against actuator faults

This work is dedicated to the robust tracking control problem for an unmanned aerial vehicle (UAV) in the presence of disturbances and actuator partial faults. Two nonlinear control strategies, which include sliding mode control (SMC) and backstepping control (BSC), are investigated and tested on a multi-input multi-output (MIMO) nonlinear quadrotor helicopter system: Qball-X4. For comparison and verifying the capability of the developed algorithms, a linear quadratic regulator (LQR) controller with integral action is also implemented and tested on a real quadrotor vehicle. The MATLAB/Simulink simulation testing results for the conditions without propeller fault, and with partial faults in all four propellers show that SMC and BSC possess strong robustness for dealing with disturbances and system uncertainties induced by faults. The real flight experiments using SMC and LQR algorithms are realized and the results further indicate that SMC has stronger robustness than LQR for dealing with disturbances and system uncertainties.

Sliding Mode Control Applied to a Qball-X4 UAV

This work is dedicated to the robust tracking problem for a UAV system with external disturbances, the Sliding Mode Control (SMC) strategy is used and tested on a Multi-Input-Multi-Output (MIMO) nonlinear quadrotor helicopter system. For comparison purpose, a LQR controller with integral action is also implemented and tested on a real quadrotor vehicle together with the SM-based controller. The results show that SMC possesses strong robustness for dealing with disturbances and system uncertainties.