Based on the actual operational parameters of a subsea multiphase pipeline, an experimental study on the internal corrosion of a subsea multiphase pipeline was conducted in a dynamic, high-temperature autoclave, which had a similar environment to an actual field environment, using the partial pressure of CO2 (P CO2 ), velocity of the corrosion medium, temperature, corrosion time, and corrosion inhibitor as variables. The results show that CO 2 resulted in severe localized corrosion and that the corrosion rate increased as the P CO2 and velocity increased; the corrosion rate first increased and then decreased with increasing temperature. The corrosion rate peaked at approximately 65˚C and then decreased continuously afterwards; the corrosion rate decreased as the duration of the experimental period increased. Under the operational conditions of the selected subsea pipeline, localized corrosion caused by CO 2 was still the primary corrosion risk. Several types of corrosion inhibitors could inhibit the occurrence of localized corrosion for a short time period; however, most corrosion inhibitors could not completely inhibit localized corrosion.

This paper proposes an extension of marginal fisher analysis (EMFA) for dimensionality reduction and analyzes some properties of both EMFA and linear discriminant analysis (LDA), and finally suggests a synthesized discriminant projection (SDP). SDP takes both global class relationship and local geometry structure into account, which maximizes the distance between marginal points and the distance between different class-centers, thus enhances classification.