In order to design efficient logistics networks, a lot of approaches are proposed especially in the OR discipline. And quite a few companies apply these methods when making decision on designing real-world logistics networks. For most conventional methods, the optimization objective is minimizing logistics cost which is defined by cargo volume handled and cost conversion factors different form cargo characteristics (e.g. product category, size). Also, capacity of distribution centers (DCs) is represented by capacity factors. Meanwhile setting values on these factors are difficult realistically because real-world logistics are realized by utilizing various operational resources, and capacity and cost should be determined by the amount of resources used for logistics operation. In this paper, we have developed a logistics network design model including resource elements of logistics to represent real-world conditions. Implementing the model, we made computational experiments varying parameters and analyzed the features of obtained logistics networks.
PID control has been widely applied in a lot of process systems. The control performance is strongly affected by tuning PID parameters, which is an extremely important issue. Many modelbased PID parameters tuning methods have been proposed, and these methods have been applying to a lot of real systems. However, in thermal processes or chemical processes, in order to calculate control parameters, experiments are required multiple times. On the other hand, the data-oriented controller design scheme has received much attention in the last few years. In this paper, a new dataoriented PID controller design scheme based on generalized predictive control (GPC) is proposed. According to the method, control parameters are obtained based on predictive values of system output, and a GPC-based PID (GPC-PID) controller is given by transforming these parameters. Moreover, in the proposed GPC-PID controller, system identification is not necessary to calculate control parameters. The effectiveness of the proposed method is evaluated by a numerical example.
In the liberalized future power market, social welfare will be maximized by demand side management. Methods of demand side management usually require exchanges of information between consumers and a utility company, but there will be strategic false reports of information. In this research we consider an energy demand network that consists of one utility company and consumers. In this network a day-ahead market is formed and we optimize the schedule of consumption using demand response under equality and inequality constraints. In order to prevent strategic false reports we apply mechanism design theory, which is discussed in economics and game theory. Our proposed integration mechanism is based on AGV mechanism, which is Bayesian incentive compatible and budget balanced. Through a numerical experiment, the effectiveness of the proposed mechanism is demonstrated under equality and inequality constraints.
More effective and efficient methodology for control education is desired, so that more people understand the control theory and that research achievements on control can be applied more easily. As an introductory part of such methodology, some basic experiences should be given. Therefore, we propose two control experiments; One is to understand phenomena such as steady state error and overshoot through tuning PID controllers. The other is to understand mechanisms of control equipments and those connections. The effectiveness of the proposed method is examined via the questionairs to students.