Noise technologies and systems for monitoring the beginning of the latent period of accidents on fixed platforms

The authors analyse the causes of frequent failures of fixed oil platforms. It has been discovered that noise in the noisy signals received by the sensors of control systems in many cases contains valuable information about the beginning of a platform's transition into the latent period of an emergency state. Considering this specific characteristic, technologies for calculating estimates of the robust correlation functions have been developed. These technologies allow one to determine the noise characteristics and form correlation matrices equivalent to the matrices of the useful signals. Based on these technologies, the authors propose systems for monitoring the beginning of the latent period of the transition of fixed platforms into an emergency state.