Makishi Nakayama

We discuss a new obstacle avoidance technique for a UAV（Unmanned Aerial Vehicle）using PSO（Particle Swarm Optimization）called optimization imitated animal foraging．In this method，the avoidance orbit is generated by the response by step change of the target position in the left and right direction and the response by the step change of the target speed in the forward direction. The magnitude of the change of the two target values is determined by the PSO based on the unique evaluation function. The evaluation function is a function consisting of collision risk based on the closest distance to the obstacle, energy based on magnitude of step change of the target value to be applied, and avoidance time based on speed change in the forward direction. Here, the calculation of the closest approach distance is obtained by performing a simulation based on the mathematical model expressing the flight trajectory of the obstacle and the own vehicle each time the evaluation function calculation of the repetitive calculation in the optimization calculation of the PSO is performed. We confirmed that this technique is effective by the numerical simulation．

ドローンタイプのUAVの高さ制御において，最適サーボ系を構築した．このサーボ系の最適ゲインは，ドローンの動特性パラメータから解析に求めることができ，最適性が保障できるゲインの範囲も明らかにした．サーボ系は単純はPID制御系として構築でき，シミュレーションと実機実験によりその有効性を確認した．

For the precise control of tension and thickness in a tandem cold mill during acceleration and deceleration, we have developed a nonlinear model including the rolling speed as a time-varying parameter. It is demonstrated from simulation results that nonlinear receding horizon control by real-time optimization results in a satisfactory performance for the nonlinear model.

制御システムの動特性のパラメータが確率パラメータで表現されるシステムのとき、出力のばらつき（分散）がもっともおさえられるようなフィードバック制御則をもとめるためのアルゴリズムを導出した。

In this paper, we consider an optimal control problem for a linear discrete time system with stochastic parameters. Whereas traditional stochastic optimal control theory only treats systems with deterministic parameters with stochastic noises, this paper focuses on systems with both stochastic parameters and stochastic noises. We derive an optimal control law for a novel cost function by which the designer can adjust the trade-off between the average and the variance of the states. Furthermore, a numerical simulation shows the effectiveness of the proposed method. © 2011 IFAC.