Ji-Hyeong Han

Preference-Based Solution Selection Algorithm for Evolutionary Multiobjective Optimization

Since multiobjective evolutionary algorithms (MOEAs) provide a set of nondominated solutions, decision making of selecting a preferred one out of them is required in real applications. However, there has been some research on MOEA in which the users preferences are incorporated for this purpose. This paper proposes preference-based solution selection algorithm (PSSA) by which user can select a preferred one out of nondominated solutions obtained by any one of MOEAs. The PSSA, which is a kind of multiple criteria decision making (MCDM) algorithm, represents users preference to multiple objectives or criteria as a degree of consideration by fuzzy measure and globally evaluates obtained solutions by fuzzy integral. The PSSA is also employed in each and every generation of evolutionary process to propose multiobjective quantum-inspired evolutionary algorithm with preference-based selection (MQEA-PS). To demonstrate the effectiveness of PSSA and MQEA-PS, computer simulations and real experiments on evolutionary multiobjective optimization for the fuzzy path planner of mobile robot are carried out. Computer simulation and experiment results show that the users preference is properly reflected in the selected solution. Moreover, MQEA-PS shows improved performance for the DTLZ problems and fuzzy path planner optimization problem compared to MQEA with dominance-based selection and other MOEAs like NSGA-II and MOPBIL.

Evolutionary Multiobjective Footstep Planning for Humanoid Robots

This paper proposes a novel evolutionary multiobjective footstep planner for humanoid robots. First, a footstep planner using a univector field navigation method is proposed to provide a command state (CS), which is to be an input of a modifiable walking pattern generator (MWPG) at each footstep. Then, the MWPG generates corresponding trajectories for every leg joint of the humanoid robot at each footstep to follow the CS. Second, a multiobjective evolutionary algorithm (MOEA) is employed to optimize the univector fields satisfying multiple objectives in navigation. Finally, a preference-based selection algorithm based on a fuzzy measure and fuzzy integral is proposed to select the preferred one out of various nondominated solutions obtained by the MOEA. The effectiveness of the proposed evolutionary multiobjective footstep planner is demonstrated through computer simulations for a simulation model of a small-sized humanoid robot, HanSaRam-VIII.

The Degree of Consideration-Based Mechanism of Thought and Its Application to Artificial Creatures for Behavior Selection

To make artificial creatures deliberately interact with their environment like living creatures, a behavior selection method mimicking living creatures thought mechanism is needed. For this purpose, there has been research based on probabilistic knowledge links between input (assumed fact) and target (behavior) symbols for reasoning. However, real intelligent creatures including human beings select a behavior based on the multi-criteria decision making process using the degree of consideration (DoC) for input symbols, i.e. will and context symbols, in their memory. In this paper, the DoC-based mechanism of thought (DoC-MoT) is proposed and applied to the behavior selection of artificial creatures. The knowledge links between input and behavior symbols are represented by the partial evaluation values of behaviors over each input symbol, and the degrees of consideration for input symbols are represented by the fuzzy measures. The proposed method selects a behavior through global evaluation by the fuzzy integral, as a multicriteria decision making process, of knowledge link strengths with respect to the fuzzy measure values. The effectiveness of the proposed behavior selection method is demonstrated by experiments carried out with a synthetic character Rity in the 3D virtual environment. The results show that the artificial creatures with various characteristics can be successfully created by the proposed DoC-MoT. Moreover, training the created artificial creatures to modify their characteristics was more efficient in the DoC-MoT than the probability-based mechanism of thought (P-MoT), both in terms of the number of parameters to be set and the amount of time consumed.

A preference-based task allocation framework for multi-robot coordination

A market-based task allocation mechanism has been widely used in multi-robot coordination. However, most of approaches consider a bid as either quality or cost, or a combination of the two. This paper proposes a preference-based task allocation framework for coordination of multiple robots. To provide preference in task allocation, four bid elements are defined: task quality, task cost, task distance and task time. Tasks can be allocated considering the relative importance of the bid elements based on the preference. The utility of a robot is calculated using the bid elements and their weighted preference. The effectiveness of the proposed framework was demonstrated through the computer simulation of a cleaning mission.

Human-robot interaction by reading human intention based on mirror-neuron system

Considering the human-robot symbiosis in coming years, robots should be able to read human intention for natural and rational interaction with human beings. This paper proposes effective human-robot interaction (HRI) by reading the human intention using cognitive architecture. Proposed intention reading algorithm is inspired by mirror-neuron system and simulation theory which are the significant parts of human mind reading skill. For human intention reading, the cognitive architecture consisting of eight modules, i.e. perception, attention, behavior mapping, behavior model, intention reading, memory, behavior selection, and actuator modules is also proposed. The effectiveness of the proposed scheme is demonstrated through computer simulations on human-robot play with two different objects, such as a ball and a toy car.

3D-position estimation for hand gesture interface using a single camera

The hand gesture interface is the state of the art technology to provide the better human-computer interaction. This paper proposes two methods to estimate the 3D-position of the hand for hand gesture interface using a single camera. By using the methods in the office environment, it shows that the camera is not restricted to a fixed position in front of the user and can be placed at any position facing the user. Also, the reliability and usefulness of the proposed methods are demonstrated by applying them to the mouse gesture recognition software system.

Multi-objective quantum-inspired evolutionary algorithm-based optimal control of two-link inverted pendulum

This paper proposes a method to generate an optimal trajectory of nonlinear dynamical system and concurrently optimize multiple performance criteria. As the dimensionality of system increases, it is difficult to find values of cost/reward function of conventional optimal controllers. In order to solve this problem, the proposed method employs iterative linear quadratic regulator and multi-objective quantum-inspired evolutionary algorithm to generate various optimal trajectories that satisfy multiple performance criteria. Fuzzy measure and fuzzy integral are also employed for global evaluation by integrating the partial evaluation of each solution over criteria with respect to users degree of consideration for each criterion. Effectiveness of the proposed method is verified by computer simulation carried out for the problem of stabilizing two-link inverted pendulum model.

Swarm intelligence-based sensor network deployment strategy

The wireless sensor network is a decentralized and self-organized system. Each sensor node in the sensor network should be intelligent enough to carry out its task of monitoring the environment. There would be numerous ways for deploying the sensor nodes in the environment. In this paper, swarm intelligence-based sensor network deployment strategy is proposed. To make a reference point for each sensor node, fuzzy integral is utilized as a multi-criteria decision making process. Three criteria, such as sensor value, crowdedness and confidence, are used for partial evaluation and the degree of consideration for each criterion is represented by fuzzy measure. Global evaluation by fuzzy integral determines the best position for each sensor node independently. To show the effectiveness of the proposed strategy, it is compared with the SPSO07-based deployment strategy through computer simulations in a simulation environment. The results show that the proposed strategy covers much wider area with sensor nodes than the SPSO07-based one.

Market-Based Multiagent Framework for Balanced Task Allocation

This paper proposes a market-based multiagent task allocation framework for allocating tasks in a balanced manner based on the energy levels of robots. In this framework, a market-based agent is designed for trading tasks considering the robot capabilities, task requirements and energy level of the robot. The framework utilizes a bid weight for distributing the tasks in a balanced manner without frequent using of particular robots. To demonstrate the effectiveness of the proposed framework, a simulation experiment was carried out for a cleaning mission consisting of collecting, carrying, sorting and disposal tasks.

Consideration about the Application of Dynamic Time Warping to Human Hands Behavior Recognition for Human-Robot Interaction

To prepare the age when humans and robots live together, robots need to understand the meaning of human behaviors for the natural and rational human-robot interaction (HRI). The robot particularly needs to recognize the human hands behavior, since humans usually express their meanings and intentions by using two hands. In this paper, the robot recognizes the human hands behavior by simulating it based on robot’s own hands behaviors set and finding the most similar one as human behavior using dynamic time warping (DTW) algorithm. To consider the effects of different variables, i.e. data normalization methods and local cost measures for DTW algorithm, this paper considers two different normalization methods and four different local cost measures and their effects are discussed. The robot successfully recognizes the eight different human hands behaviors by DTW algorithm with the chosen normalization methods and local cost measures.