Daoxiong Gong

A hybrid approach of GA and ACO for TSP

This paper proposed a hybrid approach of genetic algorithm (GA) and ant colony optimization (ACO) for the traveling salesman problem. In this approach, every chromosome of GA is at the same time an ant of ACO. Whenever GA performs the operation of crossover and mutation, the approach firstly computes the linkage strength between gene codes of parental chromosome(s) according to the pheromone matrix of ACO, and it then selects the crossover or mutation point(s) according to the linkage strength. A threshold is generated to classify the gene linkage as strong or weak, the strong linkage segments of parents are retained to offspring as far as possible. By this way, GA can avoid its useful building blocks being frequently destroyed by genetic operations. Experiments on TSPLIB validated the building block learning capability of our approach.

A new Multi-parent Recombination Genetic Algorithm

This paper proposes a new Multi-parent Recombination Genetic Algorithm, which introduces a new Fitness-Weighted Crossover (FWX) and adopts a random threshold based mechanism to determine the parent-number of multi-parent recombination. FWX is the generalization of Single-point Crossover, Two-point Crossover, Multi-point Crossover, Uniform Crossover, Arithmetic Crossover and Multi-parent Occurrence Based Scanning Crossover. It endures the fitter parent a bigger influencing factor, which is used to determine the contribution of parents to their offspring. The new GA approach guarantees the validity of the offspring by the fact that FWX is a convex combination of parents. Experiment on a suit of benchmark functions validated the advantages of our approach.

LQR control for a self-balancing unicycle robot

A self-balancing unicycle robot, which has a wheel for balancing and movement in the longitudinal plane (pitch angle) as well as a flywheel for balancing in the lateral plane (roll angle), is studied in this paper. The non-linear dynamic equations of the unicycle robot on a slope are analyzed using the Lagrangian dynamic formulation, then a linear model of the robot is derived at the equilibrium point, and 3 linear quadratic regulators (LQR) are designed to control the robot on slopes with the angle of inclination varying from -11°to 11°. Simulation results validated that the unicycle robot can achieve good performance both on level plane and on slope.

Design of a system capture human gait data in joint space

Based on the methodology of bio-inspired control of humanoid robot, human motion capture data is crucially important for improving the quality of gait as well as the ability of push-recovery of the humanoid. In order to get the human motion data that is easy to be applied to humanoid motion control, we designed a wearable human motion capture system that can capture the motion of the trunk and the lower limbs in the joint space as well as the motion of the pelvis in the Descartes space; and the forces exerted on the sole of foot are also measured so as to capture the moving of the zero moment point (ZMP) of human when he/she is walking. Both the mechanical and electronic parts of the motion capture system are introduced in detail; the advantages and the disadvantages of the system are also discussed.

Thermodynamic modeling and motion simulation of pneumatic wheeled jumping robot

To improve the jumping performance and reduce the modeling complexity, a pneumatic wheeled typed jumping robot is designed, the takeoff angle of which can be adjusted while it starts jumping. The double-acting cylinder with the power source of liquid carbon dioxide is used as jumping mechanism. The changes of the precise equations of the internal pressure in the cylinders are established and a cylinder hopping model is built. The jumping simulation and the movement of process of the cylinder are performed. Results show that the actual characteristics of movement of the cylinder agree with the simulation ones.

Dynamics modeling and real-time fault-tolerant control of a rotor aerial robot

The present micro aerial vehicle (MAV) is lack of maneuverability and security when operates in the restricted environment. To solve such problem, a rotor powered disk-type autonomous aircraft—Micro Aerial Robot(MAR) was designed, which adopts built-in blades and redundant actuators structure to enhance vehicles safety and robustness, and adopts the inner and outer two rotor systems and disk-type airframe structure, which help the MAR has better aerodynamic characters and maneuverability compared to the common MAV. The six degrees of freedom (6 DOF) dynamic model was built using the Newton - Euler method. A real-time fault-tolerant adaptive control method was designed to achieve real-time and accurate tracking control of the aircraft when the parameters associated with the input variables suffer uncertain disturbance. Simulation of the simplified mode of the aircraft were carried out in the matlab/SIMULINK environment, which verified that the aircraft had good maneuverability and robustness which is necessary when it works in the restricted environment.