Search

An improved artificial moment method is proposed for decreasing the negative effects of the robots close to their targets. In the method, a new algorithm is presented for the key companion of a robot close to its target. The algorithm will not make the robot change its key companion frequently and select the same robot as its key companion for a long duration. Second, two new algorithms are presented respectively for attractive points and attractive angles of robots. As such, a robot can decrease its negative effects on others. Finally, the existing artificial moment motion controller is improved. Simulation results indicate that the proposed method can decrease the negative effects in the system and yield better solutions in complex situations.

Mobile robot becomes more and more perfect with people research continuously, obstacle avoidance is one of the most basic function in the process of operation. At present, the robot simulation system is an important tool of researching robot. In this paper, hardware design and software design were done by NSTRSS and real-time dynamic obstacle avoidance simulation of mobile robot is completed successfully. The experimental result shows that the simulation has 3D scene, lively effect and a good operation interface.

A novel self-organizing approach to cooperative hunting by swarm robotic systems is put forward. Firstly, an individual autonomous motion planning is presented, and the cooperative hunting behaviors are mathematically described. According to decomposition of hunting behaviors, the loose-preference rule is established for the individuals to form the ideal hunting formation during the self-organizing process by the interaction between the target and individuals. Then, we employ the proposed rule to design the autonomous motion controller of the individuals. Finally, the stability of self-organizing system is analyzed by Lyapunov stability criteria. Simulations and experiments demonstrate the feasibility and effectiveness of the proposed approach to cooperative hunting by swarm robotic systems.

A zero element row augmenting path algorithm is proposed for linear assignment problems. Under the algorithm, the maximum element in an assignment y is picked firstly, where y is calculated according to the reduced matrix of a given cost matrix at the initial stage. According to each row's zero element record and starting from the row containing the picked element, a zero element row augmenting path is computed while the cost matrix is further reduced and the zero element record of each row is updated. Finally, y is adjusted according to the augmenting path so that the number of the zero elements in y can increase strictly. The above process is repeated until y is an optimal assignment. Extensive computational experiments are carried out and indicate that the proposed algorithm is even faster than the shortest augmenting algorithm.

A novel new algorithm of path planning is proposed for the robot path planning based on the local data map in the robot autonomous navigation. The algorithm is used to search for the local optimal path from the current position of the robot to the target according to the known sensor data. If the robot cannot reach the target directly, the temporary target point which the robot can reach will be set up according to the optimal path. The algorithm is effective under the complicated unknown environment and moving obstacle situation which fast searching speed, it can be adapted into the practical applications and the multi-robot harmony easily. The simulation result shows that this method can provide a better planning path by comparing with the traditional planning algorithms such as artificial potential field and wall-following

Cooperative transportation using multi-robots is a significant challenge in robotics. For the problem, each robot is required, in general, to reach a different task-point to form a transport formation, where all the task-points are determined according to the shape of the transported object and the number of robots. A method based on bottleneck-linear assignment is proposed to form complex transport formations. First, the optimal paths from each robot to all the task-points are calculated by a two-direction path algorithm, which is developed in this paper as the core of the task-points' assignment. Second, in order to optimize the travelling paths of the robots and the time taken to establish the formation, a bottleneck-linear assignment strategy is presented to assign the task-points for the robots. Finally, an improved artificial moment motion controller makes each robot move along a sub-optimal path to reach its task-point. Simulations indicate that the proposed method is feasible and efficient.

According to the problems of autonomous robotic navigation in complex environments, an opposite way path planning algorithm and a novel robot controller based on inertial property is proposed. Firstly, the opposite way path planning algorithm getting the local obstacle message through robot, according to the known obstacles a optimal path calculated which can carries the robot to get to the ultimate target. Secondly, a robot controller based on attractive force function and repulsive force function is designed. The controller can keep the robot away from the obstacle, and escape from the local minima when the robot is blocked by obstacle. The simulation results show that this method has a better optimization result than other common methods such as bug and artificial potential field.

Determinating corresponding vertexes of robots is the precondition and base for utilizing artificial moment method to control multi-robot formation. In this paper, a new criterion based on generalized assignment problems is proposed to determine corresponding vertexes of robots, and its character is discussed. Then, according to the minimum increment principle, a new method for generalized assignment problems is presented, and the program is given in MATLAB language. Finally, numerical experiments were done. Both theory and results of experiments show that the criterion is as good as the distance-optimization correspondence, and the method has many advantages, such as low time complexity, and solutions being very close to the optimal solution, so they are feasible and valid.

To control swarm-robot formations with more convenience, after defining the conception of distance-optimization correspondence, the model, principle and steps of the elimination method, an approach for distance-optimization correspondence, are described, and two basic theorems about the elimination method are also proofed. Based on above, the new model of swarm-robot systems is presented. Then a new class of artificial moment: repulsion moment of robot to robot is defined, and the new motion controller of the general robot is designed. Finally, a simulation is given and the result shows that the theories and methods proposed in this paper are feasible and valid.