| 摘要 |
A method for creating a deterministic model representing a production system such as for oil and gas includes one or more inputs and parameters that are not deterministically known, and one or more outputs. A prior probability density function is used to determine a prior uncertainty, and a measurement related to a first of the outputs is obtained, the output having a measurement uncertainty determined using a conditional probability density function conditioned on the input. A posterior probability density function on the input is determined using the prior probability density function, the measurement, and the conditional probability density function. Bayesian techniques recursively update the probability density functions representing uncertainty as new measurements become available. |
