Abstract: The approaches to make an agent generate intelligent actions in the AI field might be roughly categorized into two ways–the classical planning and situated action system. It is well known that each system have its own strength and weakness. However, each system also has its own application field. In particular, most of situated action systems do not directly deal with the logical problem. This paper first briefly mentions the novel action generator to situatedly extract a set of actions, which is likely to help to achieve the goal at the current situation in the relaxed logical space. After performing the action set, the agent should recognize the situation for deciding the next likely action set. However, since the extracted action is an approximation of the action which helps to achieve the goal, the agent could be caught into the deadlock of the problem. This paper proposes the newly developed hybrid architecture to solve the problem, which combines the novel situated action generator with the conventional planner. The empirical result in some planning domains shows that the quality of the resultant path to the goal is mostly acceptable as well as deriving the fast response time, and suggests the correlation between the structure of problems and the organization of each system which generates the action.
Abstract: In this paper, we propose a geometric modeling of
illumination on the patterned image containing etching transistor. This
image is captured by a commercial camera during the inspection of
a TFT-LCD panel. Inspection of defect is an important process in the
production of LCD panel, but the regional difference in brightness,
which has a negative effect on the inspection, is due to the uneven
illumination environment. In order to solve this problem, we present
a geometric modeling of illumination consisting of an interpolation
using the least squares method and 3D modeling using bezier surface.
Our computational time, by using the sampling method, is shorter
than the previous methods. Moreover, it can be further used to correct
brightness in every patterned image.
Abstract: Accident in spent fuel pool (SFP) of Fukushima
Daiichi Unit 4 showed the importance of continuous monitoring of the
key environmental parameters such as water temperature, water level,
and radiation level in the SFP at accident conditions. Because the SFP
water temperature is one of the key parameters indicating SFP
conditions, its behavior at accident conditions shall be understood to
prepare appropriate measures. This study estimated temporal change
in the SFP water temperature at Kori Unit 1 with 587 MWe for 1 hour
after initiation of a loss-of-pool-cooling accident. For the estimation,
ANSYS CFX 13.0 code was used. The estimation showed that the
increasing rate of the water temperature was 3.90C per hour and the
SFP water temperature could reach 1000C in 25.6 hours after the
initiation of loss-of-pool-cooling accident.
Abstract: This paper presents Cost per Equivalent Wafer Out, which we find useful in wafer fab operational cost monitoring and controlling. It removes the loading and product mix effect in the cost variance analysis. The operation heads, therefore, could immediately focus on identifying areas for cost improvement. Without this, they would have to measure the impact of the loading variance and product mix variance between actual and budgeted prior to make any decision on cost improvement. Cost per Equivalent Wafer Out, thereby, increases efficiency in wafer fab operational cost monitoring and controlling.
Abstract: Repetitive systems stand for a kind of systems that
perform a simple task on a fixed pattern repetitively, which are
widely spread in industrial fields. Hence, many researchers have been
interested in those systems, especially in the field of iterative learning
control (ILC). In this paper, we propose a finite-horizon tracking
control scheme for linear time-varying repetitive systems with uncertain
initial conditions. The scheme is derived both analytically
and numerically for state-feedback systems and only numerically for
output-feedback systems. Then, it is extended to stable systems with
input constraints. All numerical schemes are developed in the forms
of linear matrix inequalities (LMIs). A distinguished feature of the
proposed scheme from the existing iterative learning control is that
the scheme guarantees the tracking performance exactly even under
uncertain initial conditions. The simulation results demonstrate the
good performance of the proposed scheme.