칼만필터補償器를 이용한 레이다-서어보系의 最適位置制御에 관한 연구

Abstract

이 論文은 레이다-서어보계의 位置制御問題를 確率系 모델로 택하여 最適狀態推定 및 最適確率制御方法을 규명하고 이러한 制御性能을 구현할 수 있는 制御調節器를 設計하는데 목적을 두었다.

여기서 레이다-서어보系統은 적절한 가정하에 時不變線型 狀態모델로 線型化 하였으며 주어진 최종 및 초기 시간하에 이 系統을 最適化시키기 위한 制御性能 平價函數로는 狀態變數와 制御信號에 관한 二次型 형태로 취급하였다.

레이다의 角變位가 외부 確率外亂에 의하여 平衡運轉點을 이탈하였을 때 最適化 이론으로는 Stochastic Hamiltonian-Jacobi-Bellman 方程式의 最適化 原理를 도입하였으며 결과적으로 最適狀態推定 및 最適確率制御效果에 대한 결과치를 비교 검토한 결과 만족할만한 좋은 결론을 얻었다.

The objective of this paper is to estimate the optimal state and to search the control strategy of optimal stochastic position of a Radar-Servo system, considering as a stochastic model.

In the course of the analysis, the Radar-Servo dynamics are assumed to have the linear time-invariant schemes under a certain prescribed condition, and the performance measure of the state and the control effort in optimizing the system is considered to have a quadratic form when the initial and the final time is specified.

When the radar angular displacement is deviated from a stationary operating point by any external random disturbance, the optimal stochastic feedback control law invokes the stochastic Hamilton-Jacobi-Bellman equation, and suggests the realizability of the required control mechanism.

Finally the optimal state estimation and the optimal stochastic control efforts are compared to confirm the obtained results.

The objective of this paper is to estimate the optimal state and to search the control strategy of optimal stochastic position of a Radar-Servo system, considering as a stochastic model.

In the course of the analysis, the Radar-Servo dynamics are assumed to have the linear time-invariant schemes under a certain prescribed condition, and the performance measure of the state and the control effort in optimizing the system is considered to have a quadratic form when the initial and the final time is specified.

When the radar angular displacement is deviated from a stationary operating point by any external random disturbance, the optimal stochastic feedback control law invokes the stochastic Hamilton-Jacobi-Bellman equation, and suggests the realizability of the required control mechanism.

Finally the optimal state estimation and the optimal stochastic control efforts are compared to confirm the obtained results.