Nonlinear Dynamical Modeling And Comprehensive Performance Optimization Design For NW Wind Turbine Gearbox

The technology of wind turbine design is facing severe challenges with the rapid development of wind power equipment industry.The evaluation of the performance of wind turbine gearbox is reflected in two aspects,one is the smoothness of its operation,such as the vibration and noise of gear transmission system,expressed by the dynamic characteristics.Another is the service life,expressed by the reliability evaluation.Therefore,how to accurately evaluate and improve the dynamic performance and reliability of wind turbine gearbox is the key to the development of wind power transmission system.The gear transmission system of wind turbine is confronted with unique dynamics problems because of its poor working environment,complex loads and increased flexibility of structure.At the same time,the random wind loading process on the whole service life also makes the model and calculation of system reliability different from the order parts.In this dissertation,the dynamic performance and reliability index of wind turbine gearbox are studied with the NW type as the research object.The content includes Rigid-flexible coupling dynamic modeling of wind turbine gear drive,the analysis of dynamic performance indicators,and the method and process of dynamic optimum design based on nonlinear dynamics theory.The main content is as follows:(1)Dynamic modeling.An equivalent model considering the elasticity of ring gear is established including the nonlinear factors of tooth sliding friction,backlash and time-varying meshing stiffness of helical gear teeth.AR model is used to simulate the time series of wind speed for external excitation.Lagrange's cubic interpolation of PHD is used to solve the above model which not only overcomes the problem that the dynamic equations contain harmonic nonlinear terms and time-varying parameters,but also simplifies the solution process under the premise of enough accuracy.(2)Calculation and analysis of dynamic performance targets.The dynamic performance targets of wind turbine gears are studied by numerical method,and improvement measures are put forward.Then,the dynamic sensitivity of gear parameters to the performance targets are derived by direct derivative method.Finally,the influence of the gear parameters on dynamic performance is quantitatively evaluated by sensitivity analysis,which provides a theoretical basis for the parameter optimization and working performance of wind turbine gear drive system.(3)Calculation and analysis of time-varying reliability target.Considering the distribution type of wind loads is uncertain,the stress-strength interference model based on the assumption of normal distribution can't be applied to the reliability of wind turbine gear drive.For this,the dynamic reliability model for wind turbine of gear drive with uncertain distribution parameters is established firstly with the deterioration of gear contact stress and strength degradation being considered.Then,the dynamic reliability is solved by approximate expansion of the model function into standard normal distribution function based on the stochastic perturbation theory and the Edgeworth series method.In the absence of experimental data,the first four moments of the random parameters with uncertain distribution are obtained by Monte Carlo method.Meanwhile,the dynamic reliability sensitivity to gear parameters is derived by the same method and the sensitivity variation is given which effectively reflecting the influence of gear parameter changes on the system reliability of wind power gear drive,and providing an accurate theoretical basis for the optimal design of the project.(4)Proposing global optimization algorithm for mixed design variables.A hybrid optimization algorithm is presented which combining direct search method with artificial intelligence algorithm based on serial parallel strategy,while the genetic operator is introduced into the MEA method in order to enrich the diversity of the population,so as to improve the local search ability of the algorithm.The above optimization algorithm is used as the final dynamic optimization method for wind turbine gear drive.(5)Optimal design of comprehensive performance tagets for NW wind turbine gearbox in project example.In the process of dynamic optimal design,gear variables and optimization goals are determined based on parametric sensitivity analysis.BP neural network is used to fit the nonlinear mapping relationship between variables and the dynamic performance of gear drive system,which can avoid solving the differential equations repeatedly and simplify the optimization process.Finally,the dynamic optimization of the research object is carried out according to different multi-objective schemes in order to verify the feasibility and applicability of the theoretical method mentioned above for the dynamic optimal design of wind turbine gear drive.