Methodology for load estimation in mooring chains using a fuzzy genetic model

Mooring lines load prediction typically requires data from several sensors and high computational power. In this paper, a soft computing methodology is proposed to estimate the load on mooring chains based on fuzzy logic that uses only some of the natural frequencies of these systems. Two fuzzy inference systems are constructed and optimized by genetic algorithm. The performance of triangular and Gaussian membership functions in modelling the natural frequencies and their behavior in the optimization process are compared. The accuracy of this methodology is validated by experimental data obtained from a test bench. In addition, the robustness of the fuzzy inference system in dealing with uncertainties is studied using statistical parameters. Furthermore, a novel normalization method is proposed to generalize the fuzzy inference systems and eliminate the training process. The results indicate that the load on the chain can be estimated with high precision, and the proposed methodology is suitable for monitoring loads in mooring lines. •The proposed methodology requires low computational power.•Load prediction is performed using only the natural frequencies.•Normalization method eliminates the training of the fuzzy inference system.•One model can be used to estimate the load of several chains.