Optimal Sensor Placement For Predicting Vibration And Acoustic Radiation Of Cylindrical Shell Under Water

For large-scale immerged structures, such as submarines, it is important to provide relevant data of structural vibration and radiated sound field for optimistic design. To predict the real-time sound radiation under different operating conditions, large numbers of vibration sensors are usually located in the structure surface to measure the velocity, and radiated sound field can be obtained through the measured data. In engineering practice, it is difficult to implement and costs much to install too many sensors. Therefore, the research on optimal sensor placement has great engineering significance and extensive application value. In this paper, sensor placement is regarded as the main research objective to predict vibration and noise of the single cylindrical shell under water. In order to overcome the deficiency of traditional sensor placement methods, a few of improved methods are proposed and certificated, and the advantages and application scope are further discussed in this paper.Uniform sensor placement is mostly common used to predict vibration and sound radiation of cylindrical shell structures. And random-uniform method in this paper can be seen as an extension of the above one. Therefore, it inherits the advantages of uniform method, that sensors are well distributed and covered all-around. In the same time, the improved method can avoid the subjectivity and blindness when using uniform method to pick sensors. The second optimal sensor placement method is put forward based on proportion-coefficient effective independence, which introduces proportion-coefficient to reflect the correlation of sensors’ mode information. And this method can not only solve the problem of mode information redundancy effectively, but also predicts the vibration and sound radiation accurately. By combining QR decomposition and genetic algorithm, the third optimal sensor placement method can obtain high accuracy of forecast results and save much time of search algorithm.According to engineering practice, the influence of different factors’ on forecast results are discussed, such as some parts of the structure can’t be used to install sensors, one or two installed sensors are deviated from desired position or become invalid and so on. Through comparing the forecast results in different conditions, the engineering applicability of every improved sensor placement method are listed.Finally, based on MATLAB GUI, a system is designed for optimal sensor placement and predicting vibration and sound radiation of cylindrical shell structures, and it is convenience and easy for users to operate.