Optimal strategies for monitoring a variable subjected to random changes

摘要

A criterion is presented for selecting the number of samples and intersample interval time in the monitoring of a variable subjected to random fluctuations. The criterion consists in minimizing both the covariance error of the estimate of the variable and the cost of the monitoring program. By solving this two-objective minimization problem, the Pareto set of optimal solutions is obtained. Finally, an example based on actual data is given.