In the large scale material processing factory such as iron-and-steel works, aluminum plate rolling or copper stripping processes, the jobs of the same type of processing method are grouped to one operational lot from the viewpoint of productivity and quality based on the operating conditions specific to the facility and the priority which is given to the processing jobs in accordance with particular situations. To support production planning or considering operation rules and capital investment policy in such processes, we have already proposed a simulation model, which has hierarchical queue structure to express various lot making operations flexibly. In addition we newly implemented an estimation mechanism of the change in the number of jobs included in the lot to reduce unnecessary staying time of jobs caused by further lot size enlargement. In this paper the outline and features of the proposed model are described and the effectiveness of the model is confirmed by numerical experiments. Finally verification examples using actual production system data are introduced.
A ship is turned by steering. Then the ship turns with constant angular velocity against the steering angle. However, the turning angular velocity is varied by ship’s speed. Furthermore, the ship’s turning performance is based on the speed against the water. This study designs steering control system to keep the constant turning angular velocity on the sea a tidal current. First, we inspect the variation of the turning angular velocity in regard of some ship’s speed. Second, the ship’s maneuverability is identified, which is expressed as Auto-Regressive eXogeneous (ARX) model, by each steering angle and each ship’s speed. Then the angular velocity of turning is controlled by applying Generalized Minimum Variance Control (GMVC) as one of predictive controls. Finally, the gain of GMVC controller is scheduled with the tidal current effect.
Many researchers have tried to clarify the mechanism of growth of complex networks by developing network generation models like WS model proposed by Watts and Strogatz and BA model proposed by Barabási and Albert. A new network generation model is proposed in this study. It gives a sense of characteristics or interest to each model in the network. In our proposed model, a node selects a target node to make a directed link based on the latent topic. This study examines the features of the networks generated by our model through computer simulation. In the examination, we calculate the average path length, clustering coefficient and power exponent, which are representative evaluation indices of network, and check whether it satisfies the property of complex networks. We also qualitatively confirm whether our model generates networks with communities of a certain size by visualizing them.
In this paper, a distributed cooperative full-state observer is considered for discrete-time linear systems, where it utilizes a consensus mechanism which enables local observers to exchange their estimates in an appropriate interval. The behavior of its estimation error is derived and a necessary and sufficient condition such that the error asymptotically converges to zero is presented. The selection of appropriate gains of the observer is also investigated. Then it is shown that there exists an information exchange interval such that the estimation error asymptotically converges to zero if the overall system is detectable and the structure of information exchange graph between the local observers is connected and undirected. A numerical example is presented that the proposed method shows the desired behavior.