| 摘要 |
PURPOSE:To obtain an effective state variable vector even in case the mean value of disturbance is not equal to zero and also unknown, by transmitting said variable vector in parallel through a proportional feedback a time average feedback element and a trend feedback element respectively. CONSTITUTION:An optimum control input vector (20) is obtained from a state equation (1), where (t): time, t0: initial time point, (x): n-dimensional state vector, (u): r-dimensional control input vector, (w): m-dimensional disturbance and a disturbance vector w(t) supposed as the normal white noise respectively. The 1st, 2nd and 3rd terms of the right member of the equation 20 are attained by the proportional feedback of the state variable vector, the time average feedback of the state variable vector and the trend feedback of the state variable vector respectively. Then the sum of these vectors of feedbacks is obtained and used as a control input vector to attain the equation 20. When an optimum regulator is obtained based on the equation 20, P, I, D and K are decided by equations 21, 22, 23 and 24 respectively.
|
