This paper presents a data assimilation technique for social agent-based simulation to fit real world data automatically by a reinforcement learning method. We used the hidden Markov model in order to estimate the states of the system during the reinforcement learning. The proposed method can improve simulation models of the social agent-based simulation incrementally when new real data are available without total optimization. In order to show the feasibility, we applied the proposed method to a housing market problem with real Korean housing market data.

Though agent based models are used in many domains, the usage have been either very abstract model for conceptual experiments or very detailed models with huge engineering efforts in their modeling and calibration. One reason of this limited usage comes from the difficulties in calibrating and validating the model with observed data because the models are very generative in its nature with many hand-picked parameters. This paper presents a noble framework of augmenting machine learning techniques to agent-based models for better calibration and validation. The framework identifies periods of deviation between the simulation and the observation with hierarchical Dirichlet process hidden Markov Model, or HDP-HMM, and the framework automatically calibrates the temporal macro parameters by searching parameter spaces with more likelihoods of validation. After iterations of this framework, our experiments demonstrated sucessful validations on a hypothestical simple segregation model as well as a real world real estate model. This framework is generally usable in any agent based models with temporal macro parameters, which could be true in many existing models.

In this paper, we present the design and implementation of an UPnP based media distribution system supporting priority-based Quality-of-Service (QoS) in a home network that converges data and multimedia traffic. We adopt two UPnP based standards, the UPnP AV architecture and the UPnP QoS architecture, for multimedia networking and QoS management among networked devices, respectively. Additionally, the IEEE 802.1p/Q is adopted as a priority-based QoS scheme for an Ethernet-based home network. By combining two adopted architectures with the 802.1p/Q QoS scheme and integrating them into a home gateway and multimedia devices, our system offers not only interoperability of media distribution among different networked devices but also easy to use management of QoS for multimedia.

Many practical control problems for the complex, uncertain or large-scale plants, need to simultaneously achieve a number of objectives, which may conflict or compete with each other. If the conventional optimization methods are applied to solve these control problems, the solution process may be time-consuming and the resulting solution would often lose its original meaning of optimality. Nevertheless, the human operators usually perform satisfactory results based on their qualitative and heuristic knowledge. In this paper, we investigate the control strategies of the human operators, and propose a fuzzy model-based multi-objective satisfactory controller. We also apply it to the automatic train operation (ATO) system for the magnetically levitated vehicles (MAGLEV).

In this presentation we propose an algorithm for the adaptive gait control by which a quadruped walking robot can walk against external disturbances. The proposed gait mechanism makes it possible for a quadruped walking robot to change its gait pattern in response to external disturbances from various external forces acting on the body of the walking robot. More specifically, we propose a new criterion for the stability margin of a walking robot under unknown external forces by using the effective mass center based on the zero moment point. And for the solution of the adaptive gait algorithm against external disturbances, a method of altitude control and reflexive direction control is suggested. Based on the algorithmic search method, an optimal stride of the quadruped walking robot that walks from any leg configuration to the standard configuration is also proposed. To verify the efficiency of the proposed approach, some simulation results are provided.