In the smart grid system, dynamic pricing can be an efficient tool for the service provider which enables efficient and automated management of the grid. However, in practice, the lack of information about the customers' time-varying load demand and energy consumption patterns and the volatility of electricity price in the wholesale market make the implementation of dynamic pricing highly challenging. In this paper, we study a dynamic pricing problem in the smart grid system where the service provider decides the electricity price in the retail market. In order to overcome the challenges in implementing dynamic pricing, we develop a reinforcement learning algorithm. To resolve the drawbacks of the conventional reinforcement learning algorithm such as high computational complexity and low convergence speed, we propose an approximate state definition and adopt virtual experience. Numerical results show that the proposed reinforcement learning algorithm can effectively work without a priori information of the system dynamics.

No abstract

No abstract

As semiconductor features shrink in size and pitch, the extreme control of CD uniformity, MTT and image placement is needed for mask fabrication with e-beam lithography. Among the many sources of CD and image placement error, the error resulting from e-beam mask writer becomes more important than before. CD and positioning error by e-beam mask writer is mainly related to the imperfection of e-beam deflection accuracy in optic system and the charging and contamination of column. To avoid these errors, the e-beam mask writer should be designed taking into account for these effects. However, the writing speed is considered for machine design with the highest priority, because the e-beam shot count is increased rapidly due to design shrink and aggressive OPC. The increment of shot count can make the pattern shift problem due to statistical issue resulting from e-beam deflection error and the total shot count in layout. And it affects the quality of CD and image placement too.In this report, the statistical approach on CD and image placement error caused by e-beam shot position error is presented. It is estimated for various writing conditions including the intrinsic e-beam positioning error of VSB writer. From the simulation study, the required e-beam shot position accuracy to avoid pattern shift problem in 22nm node and beyond is estimated taking into account for total shot count. And the required local CD uniformity is calculated for various e-beam writing conditions. The image placement error is also simulated for various conditions including e-beam writing field position error. Consequently, the requirements for the future e-beam mask writer and the writing conditions are discussed. And in terms of e-beam shot noise, LER caused by exposure dose and shot position error is studied for future e-beam mask writing for 22nm node and beyond.

No abstract