OBSERVED VALUE ANALYZING METHOD

摘要

PROBLEM TO BE SOLVED: To easily calculate a square root matrix of the inverse matrix of an error covariance matrix by finding the integral value of an observed value at certain time by correcting its observed momentary value and the integral value of an observed value which is a certain time before as specified. SOLUTION: An integral processing part 2 integrates the observed value at certain time which is imputted from am input terminal 1 by multiplying its momentary value and the integral value of the observed value which is the certain time before by weightαandξbased uponα=2/((T/Δt)+1) andξ=1-d(T is integration time,Δt is observation cycle of observed momentary value) and adds the results together, thereby improving the S/N. Then, an observed value information calculation part 4 estimates error variance of an observed value time series at each time and an inverse matrix square root estimation part 5 estimates the square root matrix of the inverse matrix of the error covariance matrix according to the obtained error variance and integral parameters. Further a state quantity estimation part 3 estimates the state quantity of the observed system by a nonlinear method of least squares by utilizing the square root matrix.