One of the most important effects of corrosion in concrete structures is the reduction of the mechanical properties of steel-reinforcement. This is why structural steels with corrosion resistance characteristics are commonly used to extend the service life of the concrete structure, since it is not appropriate to have a structure that only satisfies characteristics of mechanical resistance if a suitable durability is not guaranteed. The ability to evaluate and control the phenomena of electrochemical corrosion by sophisticated techniques remains a research topic of great interest at worldwide level.

So, in this research work is intended to evaluate the electrochemical behavior of the reinforcing steel embedded in concrete by using the electrochemical noise EN technique with artificial intelligence AI theory. To carry out the electrochemical evaluation, specimens of reinforced concrete were manufactured with standard steel rods and different c/w ratio, and exposing in a tropical marine atmosphere for a several time. EN is performed using stochastic oscillation measurements of the corrosion potential E_corr and current I of two identical working electrodes with respect to a reference electrode. In this type of technique plots of I and E vs. time are obtained. The EN data will be analyzed using statistical and mathematical methods such as variance, standard deviation, rapid Fourier transform (FFT), Wavelet transform (WT) and also a neuronal network has been obtained to predict the corrosion damage on reinforced-steel concrete.