Study Of Multi-Objective Optimization Design Of Large Scale Wind Turbine Blade

The blade is the key component of large scale wind turbine, whose multi-discipline and multi-criteria integrated optimization design has always been emphasized. Large scale wind turbine blade design is a complicated multi-objective optimization problem. At present, an trial-and-error design process as “aerodynamic design- structural design- verification- modification” is usually adopted in wind industry,where the blade aerodynamic design and structural design are separated,and the coordination of the multi-variables, multi-objectives and multi-constraints relies on engineering experience; so much so that a feasible solution instead of an optimal solution can only be obtained and it may not match the wind turbine system the best. Therefore by means of improving the design method to realize the blade optimal design and the matching ability with the whole machine, it is of vital significance for the development of wind energy industry.To overcome the difficulties in blade design method, the coupling evolutionary algorithm and priori vector method is investigated,based on which a constrained multi-objective evolutionary algorithm is established and presented in this dissertation. The blade unsteady aerodynamic calculation, structural modelling and wind turbine structural dynamic calculation are then combined into this coupled method, settting up a blade aerodynamic/structural integrated optimization design method for large scale wind turbine. This method can be used to high-efficiently handle blade optimization design problems of complex muti-objectives, multi-variables and multi-constraints with good robustness. some main efforts have been made in this thesis as follows:For the urgent need of high performance optimization algorithm in multi-objective optimization design, Utopia reference vector system is constructed in the framework of the multi-objective evolutionary algorithms. With the analyse of the distribution characteristics of the optimal solution set for the current population, key factors such as anchor points, Utopia plane and Utopia vector, etc., are obtained. This resolves the traditional projection problem in the perpheral region and elaborately couples the diversity maintaining strategy in the classic algorithms and evolutionary algorithm. The polymerization fitness distribution and population association mechanism are constructed. A series of matching efficient operators including strength Pareto non-dominating delamination strategy, Latin hypercube generation of initial population, fields choice, SBX crossover and polynomial mutation are chosen. An high-efficiency multi-objective evolutionary algorithm, called single reference system multi-objective evolutionary algorithm(SRS-MOEA), is then developed based on vector guide. The analysis of a series of test functions shows that the SRS-MOEA can obtain uniform distribution of a complete Pareto optimal solution set through a single process for simplex and continuous front problems, maintaining excellently the population diversity and enhancing significantly the optimization efficiency.For the complex front optimization problem that SRS-MOEA cannot solved, a double dimensionless Utopia reference vector system is set up and a dynamic coupling strategy between the two vectors are established. New aggregation fitness, strength Pareto non-dominating delamination and population association mechanisms are included. a new method is developed and combining that are built independently. The efficient operators with the abilities of local exploration and development ability are chosen and integrated. The constraint treatment mechanism is also included. Based on these studies, an efficient multi-objective evolutionary algorithm, called multiple reference system multi-objective evolutionary algorithm(MRS-MOEA), for complex front and constrained optimization problems is developed. The analysis of a series of test functions show that MRS-MOEA has a very strong ability and an excellent performance in dealing with complex front, multi constraints, small population and high dimensional optimization problems. This will provide a reliable high-performance optimization algorithm for blade design method.Large-scale wind turbine blade extreme load calculation method and platform have been self-developed. Wind machine aeroelastic calculation, design load case generation and extreme load analysis are integrated into large wind turbine design auxiliary software package named Hawt Cad which is developed independently by the workteam. According to the rules and regulations of GL(2010), blade extreme load calculation method and application platform suitable for large-scale parallel computing have been established. The calculation accuracy and reliability are validated by extreme load calculation of a 2 MW blade as an example.A high-performance multi-objective optimization design method for large-scale wind turbine has been developed. Blade parameterization modeling method is established. The mathematical models for the primary design goals and key constraint conditions are also established. Combining the multi-objective evolutionary algorithm with blade extreme load calculation method, the optimization design method of blade is developed. based on this method, two-, three- and four- objective optimization design for a large 1.5 MW wind turbine blade are carried out. The optimization results show that the multi-objective optimization design of the large wind turbine blade is a complex continuous front problem. This design method can efficiently achieve convergence, and obtain a large and uniform distribution of integrated optimal solutions in a single process. Finally, the 1.5 MW blade design achieves optimal matching between aerodynamics and structure with its high aerodynamic efficiency and good mechanical properties validated by wind tunnel test and finite element verification, respectively.