Research On Sensor Placement Method For Damage Diagnosis Of Wind Turbine Blade

As one of the key and basic components of the whole wind turbine generator, wind turbine blade has always been regarded as the lifeblood of wind power business by industry personnel and it's health status is the premise of the safe operation of the wind turbine. The wind turbine generator always works under alternating load and in the wild environment of dust, rain and snow, so structural damage of blade is easy to happen. Although such damage will not affect the normal use of blade immediately, the partial damage may expand or enlarge with accumulated damage, making potential threats to the safe operation of whole wind turbine generator. Therefore, the optimal sensor placement which is a very important component of blade structural damage detection system, has become an inevitable direction of blade research. It can extremely cut detecting costs and improve the stability of the detecting system to get mode data as much as possible with as little as possible of the sensor. In this dissertation, the finite element simulation and physical experiment are used to study the optimal sensor placement of the wind turbine blade. The main work is as follows:Firstly, based on analyzed structure and form of blade, the accurate wind turbine blade model is established, and modal analysis of the model is carried out. By using ANSYS, combined with the basic parameters of the blade, the process of the model building of 1.2MW large blade is introduced in detail, and the selection of the material properties and unit type are discussed. The first five natural frequencies and their vibration modes are acquired by structural modal analysis technique.Secondly, aiming at the deficiency of traditional sensor placement algorithm applied to blade damage diagnosis, a combinational algorithm for sensor placement which integrates the advantages of effective independence method, kinetic energy method and modal assurance criterion is proposed. The combinational algorithm and two traditional algorithms are applied to detecting system of 1.2 MW blade to obtain three layout schemes. By comparison, the combinational algorithm has better mode observability.Thirdly, the obtained optimal sensor placement schemes through using three algorithms is applied to the 15 kW blade by simulation. A laboratory bench for blade dynamic performance test is set up according to the three schemes. The vibration signal of the blade is picked up by the acceleration sensor and the on-line data acquisition instrument. By using the power spectrum analysis, the measured vibration shapes and inherent frequencies of 15 kW blade are determined.Finally, based on comparing the practical shapes and inherent frequencies with those calculated by ANSYS, it is obtained that combinational algorithm is more close to the simulation results. Therefore, it is verified to effective for using the combinational algorithm proposed in this dissertation for blade optimal sensor placement, In turn, the model of 15 kW blade is also proved to be accurate.