Jeong-Jung Kim

An Artificial Pneumatic Muscle Control Method on the Limited Space

Pneumatic muscle has many advantages such as elasticity, high power and structural similarity to a living things muscle. There have been many researches to control robot actuated by pneumatic muscles, but conventional theories are hard to apply on real robot plants because of their assumptions and disregards of pneumatic muscles physical aspects like size of pneumatic muscle and its controller. Here, the new method for saving space which is occupied by many controllers to operate robot actuated by pneumatic muscles is proposed. Actually there is easy way to control pneumatic muscle using the commercial proportional pressure regulator, but its size is not suitable to be embedded on stand alone robot. So, new method using the pressure switches of compact size and encoders is suggested. This new method is tested on a robot link with ball joints, actuated by four pneumatic muscles

Improved Ant Colony Optimization algorithm by potential field concept for optimal path planning

In this paper, an improved ant colony optimization (ACO) algorithm is proposed to solve path planning problems. These problems are to find a collision-free and optimal path from a start point to a goal point in environment of known obstacles. There are many ACO algorithm for path planning. However, it take a lot of time to get the solution and it is not to easy to obtain the optimal path every time. It is also difficult to apply to the complex and big size maps. Therefore, we study to solve these problems using the ACO algorithm improved by potential field scheme. We also propose that control parameters of the ACO algorithm are changed to converge into the optimal solution rapidly when a certain number of iterations have been reached. To improve the performance of ACO algorithm, we use a ranking selection method for pheromone update. In the simulation, we apply the proposed ACO algorithm to general path planning problems. At the last, we compare the performance with the conventional ACO algorithm.

Improved Ant Colony Optimization algorithm by path crossover for optimal path planning

In this paper, an improved Ant Colony Optimization (ACO) algorithm is proposed to solve path planning problems. These problems are to find a collision-free and optimal path from a start point to a goal point in environment of known obstacles. There are many ACO algorithms for path planning. However, it take a lot of time to get the solution and it is not to easy to obtain the optimal path every time. It is also difficult to apply to the complex and big size maps. Therefore, we study to solve these problems using the ACO algorithm improved by the path crossover scheme. The path crossover scheme is two-point crossover paths found by ants. The best path is stored and is compared with new path every time. The path crossover scheme is used at this time. When the two parts compared and exchanged, the better part updates the best path. We also propose that the pheromone update rule is modified as compared with previous our paper.

Fast human detection using Gaussian Particle Swarm Optimization

Human detection is a challenging task in many fields because it is difficult to detect humans due to their varying appearance and posture. The evaluation speed of the method is important as well as its accuracy. In this paper, we propose a novel method using Gaussian Particle Swarm Optimization (Gaussian-PSO) for human detection with the Histograms of Oriented Gradients (HOG) feature to achieve a fast and accurate performance. Keeping the robustness of HOG feature on human detection, we raise the process speed in detection process so that it can be used for real-time applications. These advantages are given by a simple process which needs only one linear-SVM classifier with HOG features and Gaussian-PSO procedure.

SDAT: Simultaneous detection and tracking of humans using Particle Swarm Optimization

Human detection is a challenging task in many fields because it is difficult to detect humans due to their variable appearance and posture. Furthermore, it is also hard to track the detected human because of their dynamic and unpredictable behavior. The evaluation speed of method also important as well as its accuracy. In this paper, we propose Simultaneous Detection and Tracking (SDAT) method using Gaussian Particle Swarm Optimization (Gaussian-PSO) for human detection with the Histograms of Oriented Gradients (HOG) features to achieve a fast and accurate performance. Keeping the robustness of HOG features on human detection, we raise the process speed in detection and tracking so that it can be used for real-time applications. These advantages are given by a simple process which needs just one linear-SVM classifier with HOG features and Gaussian-PSO procedure for the both of detection and tracking.

Falling avoidance of biped robot using state classification

This paper introduce a state classification method for detecting falling of biped robot. The method uses a support vector machine (SVM) to classify the state. The input vector for the SVM are a magnitude of acceleration, a position of center of pressure (CoP) in x and z axis, and tilt angles of torso relative to x and z axis. The input vector is based on sensor data that is measured from accelerometer and force sensing resistor (FSR) sensor. Training of the classifier is done in off-line and the trained classifier is used to classify the state of the biped robot in on-line. The method was verified in a 3D dynamics simulator and showed it could classify falling state within 0.01 second.

Speeded-up Cuckoo Search using Opposition-Based Learning

For several decades, swarm intelligence (SI), emergent collective intelligence of groups of simple agents, has been applied to diverse research areas including optimization problems. Particle swarm optimization, ant colony optimization, artificial bee colony algorithm are well-known examples, and many variants are proposed so far. Recently proposed cuckoo search is also one class of SI. It mimics behaviors of cuckoo: intraspecific brood parasitism, cooperative breeding, and nest takeover. From the previous studies, it has quite a potential, so that it could outperform existing algorithms such as PSO. However, with respect to the convergence, CS shows slow performance. In this paper, we combine opposition-based learning (OBL) with CS, so that the convergence speed of CS becomes faster, not deteriorating the search ability of the algorithm. Through the simulation, the results indicate that the proposed algorithm outperforms the original algorithm not only in terms of convergence speed but also in terms of solution accuracy and success rate.

Unified Detection and Tracking of Humans Using Gaussian Particle Swarm Optimization

1 ) 1 KAIST Abstract: Human detection is a challenging task in many fields because it is difficult to detect humans due to their variable appearance and posture. Furthermore, it is also hard to track the detected human because of their dynamic and unpredictable behavior. The evaluation speed of method is also important as well as its accuracy. In this paper, we propose unified detection and tracking method for humans using Gaussian-PSO (Gaussian Particle Swarm Optimization) with the HOG (Histograms of Oriented Gradients) features to achieve a fast and accurate performance. Keeping the robustness of HOG features on human detection, we raise the process speed in detection and tracking so that it can be used for real-time applications. These advantages are given by a simple process which needs just one linear-SVM classifier with HOG features and Gaussian-PSO procedure for the both of detection and tracking.

Modified A* algorithm for outdoor environments with risk and velocity map

A* algorithm can generate the optimized unique path if the cost function is well designed based on the given information. Especially for global path planning (GPP), A* is one of best solutions because GPP itself assume that a prior information. In this paper, classic A* algorithm is modified for outdoor military robot with risk and velocity map. For this purpose, we design velocity map using digital elevation map (DEM) and the cost function for considering risk and velocity map.

Experience repository based Particle Swarm Optimization and its application to biped robot walking

In this paper, experience repository based particle swarm optimization (ERPSO) is suggested for effectively applying particle swarm optimization (PSO) to real life problems. The ERPSO uses a concept experience repository to store previous position and fitness of particles to accelerate convergence speed of PSO. The proposed method was compared with PSO variants in a three dimensional dynamic simulator for the bipedal walking. The ERPSO found the best fitness value and central pattern generator parameters that could produce a walking of a biped robot. And ERPSO has fast convergence property which reduces the evaluation of fitness of parameters in a real environment.