No abstract

Background: Due to the complexity of ocean environmental loading models, together with the nonlinearity and empirical parameters involved in hydrodynamic numerical modeling and model testing, many uncertainties and challenges still exist in the design and operation of platforms built to float at sea. On-site prototype measurements carried out on actual floating structures provide a valid strategy for obtaining accurate environmental loading parameters and floater motion responses. Problem definition: A prototype monitoring system has been built as part of a joint industrial project in the South China Sea. A complete set of long-term ocean environment loading parameters and structural dynamic motion responses has been gathered for the period from 2012 to the present. Several advanced techniques, such as the independent remote monitoring technique (IRMT), the integrated continuous measurement technique and the standalone underwater measurement technique were established to enhance the reliability of data collection even during extreme typhoon conditions when there was no power. Solution approach and findings: this paper analyzes the dynamic motion characteristics of the platform structure based on the monitoring data. The relationship between the measured wave spectrum and the JONSWAP spectrum is discussed. The spectral shape parameters of the JONSWAP spectrum for the South China Sea, as derived from the monitoring data, are discussed. The dynamic motions of the platform structure are analyzed based on artificial neural networks (ANN) using data from a typical monitored typhoon. The numerical modeling used in this research is constructed to perform the identification analysis of the platform parameters using radius basic function (RBF) and hydrodynamic results produced by ANSYS-AQWA. This research selects five main geometric parameters related to the platform design. Mass, moments of inertia of three rotation degrees, and the position of the center of gravity (COG) are selected as the optimization objectives. The mean values of surge and pitch and standard deviations of roll and pitch are treated as the input parameters. Modeling verifications show that the present ANN-based method performs well in obtaining the optimal platform parameters. The maximum error between the simulated and monitored results in terms of the measurement of the roll, pitch, surge and sway motions fall within 5%. The model of the monitored platform could be further updated; it could be made capable of performing the performance assessments of the dynamic characteristics in extreme and/or harsh environmental conditions.

At present, the complexities of distributive characteristics in temporal and spatial domain of ocean environmental loading contribute to the difficulties of the fatigue life estimations of marine structures. In shallow water, soft yoke mooring system (SYMS) is considered to be the best mooring system, and has been widely used in oil development in the Bohai Bay and the Gulf of Mexico. Soft yoke mooring system establishes the mooring functions via the multi-dynamic mechanism of thirteen hinge joints. The accuracy of fatigue life of the hinge joints is important to ensure the safety of mooring system. The damage failure of hinge joints would cause great financial loss. In 2012, Dalian University of Technology set up a full coupled proto-type monitoring system which consisted of the four sub-monitoring systems, that is, ocean environmental parameters sub-system including wind, current and wave factor, motions and attitudes of the FPSO including six degree freedom of vessel motions, motions and the mooring force monitoring system of the mooring leg. The massive monitoring information is obtained by the integrated software with continuous. The present paper proposes a real-time fatigue life prediction method of upper hinge joint of SYMS based on the prototype monitoring technique. The friction parameter of hinge joints contact surface is increased in long-term service and reduced by adding lubricant. In the SYMS design phase, there is no effective analysis of the repeated friction parameter changes. The variations of friction coefficients caused by long-term cycle stress and maintenance are considered in the fatigue calculation. The stress distribution of hinge joints under design parameter is carried out by using ABAQUS. Through calculation and comparison, the equivalent stress and fatigue damage variable of KPA (Key Process Area, large deformation units and easy wear area) units in the condition of the friction coefficient is 0.15 (design parameter) and 0.95. We found that the friction coefficient change due to long-term service will speed up the fatigue failure of the hinge joints. The relationship between friction coefficients and KPA regional stress of mooring legs swinging angle are established through the finite element simulation. Through prototype monitoring software analysis the marine environment loading, structural response and KPA regional stress information, the abrasion of the hinge node and fatigue damage variable Dθμ can be real-time predicted. The present fatigue life analysis method based on monitoring technique exhibits good advantages and research value for the fatigue life estimation of offshore structure subject to wave induced motions.

With the widely application of cluster analysis, the number of clusters is gradually increasing, as is the difficulty in selecting the judgment indicators of cluster numbers. Also, small clusters are crucial to discovering the extreme characteristics of data samples, but current clustering algorithms focus mainly on analyzing large clusters. In this paper, a bidirectional clustering algorithm based on local density (BCALoD) is proposed. BCALoD establishes the connection between data points based on local density, can automatically determine the number of clusters, is more sensitive to small clusters, and can reduce the adjusted parameters to a minimum. On the basis of the robustness of cluster number to noise, a denoising method suitable for BCALoD is proposed. Different cutoff distance and cutoff density are assigned to each data cluster, which results in improved clustering performance. Clustering ability of BCALoD is verified by randomly generated datasets and city light satellite images.