Probabilistic Dynamic Optimization Design For Support Structure Of Offshore Wind Turbines

Offshore wind turbine is the key equipment in offshore wind energydevelopment. It has complex structures and locates in severe seaenvironments, which contains random factors that make the installation,transportation and safety face many challenges.The offshore wind turbine fixed support structure is the transitionbetween the legs under the mud line and the tower above the water level. Andthe tower supports the turbine and blades. These supports are the key parts ofthe offshore wind power system. They support the whole weight of theturbine and transfer loads, and bear winds, waves, currents and earthquakeetc. With almost all wind turbine parts mounted on the tower, tower collapseoften resulting in devastating losses for the entire wind turbine. Therefore, theresearch of offshore wind turbine support structure is the key point inoffshore wind power development and need to be in-depth study.The fixed support structure of offshore wind turbine is mainly facingtwo problems: one is the cost problem, and the other one is the safety andreliability problems. On the premise of ensuring the safety and reliability ofthe structure, make an optimization design to lower the cost. The seaenvironment is complicated, and at the same time the material, size, loadingstate of support structure have randomness, both lead to the uncertainty of theresults. So the traditional deterministic optimization in the practicalengineering lacks enough reliability. This study adopts stochasticoptimization design to make the offshore wind turbine support structure saferand more reliable.Research is on the stochastic dynamic optimization problem for offshorewind turbines under dynamic environmental loads. The main contents areas follows:1. Support structure dynamic analysis in marine environment Dynamic loads on the wind turbine support structure in the marineenvironment are studied. Dynamic environmental loads include turbine loadsto tower top, wind loads on tower surface, wave and flow loads on the tripodsupport structure, as well as seismic loads. They are applied on the finiteelement model and the dynamic analysis are performed: modal analysis todetermine the natural frequencies of support structure to avoid the bladepassing frequency and turbine rotor frequency; The time domain analysis togive the von mises stress and the displacement responses of the offshore windturbine support structure in regular work and emergency. The displacementresponses under earthquake are also considered. All of these responses areregard as the constraint conditions in the optimization. At the same time, thestructure dynamic analysis in extreme turbulence cases is applied. They are inthree typical regions and two metocean conditions about50-year and100-year return, the turbine are under HSS brake shutdown and parking，there is a total of12cases in which stress and displacement responses areanalyzed and compared.2. the optimization model of tripod support structureOrthogonal Arrays is applied as one of experimental design method toreduce design parameters. Design parameters are sorted by contributions todynamic responses.4parameters which are thicknesses of column and legand diameters of brace and leg are kept after sorting. Optimal Latinhypercube design is applied to construct approximation models, here isKriging model, which show relationships between variables and responses asapproximation models. Through the approximation models, the computationcosts spent on the dynamic analysis using FEA and in stochastic optimizationprocess have been greatly reduced. It provides stable and efficient iterativeenvironment for the structure optimization.3. Stochastic dynamic optimization for the tripod support structureThe stochastic dynamic optimization for the tripod support structure isdiscussed. The numerical optimization algorithm that external penaltyalgorithm and the global search method that multiple island genetic algorithmare combined to optimize the tripod support structure. A single-loop methodis used to reduce the optimization iterations and significantly improve thecomputational efficiency. Optimized results meet a lower cost and a better reliability than before. It scientifically takes account of the economy andsafety of the support structure, overcome the problem of shortage ofreliability in the traditional support structure deterministic optimizationdesign.In this research, the finite element model of tripod support structure ismodeling to get the structure mechanical properties and dynamic responses.Introduce large quality method to apply seismic loads which reasonableconsidering seismic responses of the offshore wind turbine support structure.Adopt Kriging model instead of finite element model greatly reduceoptimization time and improve the stability of the optimization. Study on thestochastic dynamic optimization design fully considers uncertainties causedby random factors like the structure sizes, materials, environment loads andso on. Combination of the reliability analysis and structure optimizationmeets the requests of both the economic and safety which has certainreference significance in engineering design.