The design of systems important to the safety of nuclear power plants (NPPs) is based on compliance with regulatory documents’ requirements and ensuring the object’s reliable operation throughout its entire life cycle. One of the aspects of increased requirements for the reliable and safe operation of NPP process control systems is perfecting human-operator information decision support systems. In the paper, an approach is considered in which the operator support information system kernel is a flexible modeling program complex, which, depending on the tasks being solved, can be used as a digital twin or a digital shadow. The digital twin furnishes the past state of the product, the current state, and probable future states, which is especially significant for its use. The paper also discusses methods for assessing the proximity of a physical object and its digital twin based on a consistent and normalized measure of dependence of random vectors.