Uncertainty analysis of galloping based piezoelectric energy harvester system using polynomial neural network

Abstract

This paper deals with the impact of uncertain input parameters on the electrical power generation of galloping-based piezoelectric energy harvester (GPEH). A distributed parameter model for the system is derived and solved by using Newmark beta numerical integration technique. Nonlinear systems tend to behave in a completely different manner in response to a slight change in input parameters. Due to the complex manufacturing process and various technical defects, randomness in system properties is inevitable. Owing to the presence of randomness within the system parameters, the actual power output differs from the expected one. Therefore, stochastic analysis is performed considering uncertainty in aerodynamic, mechanical, and electrical parameters. A polynomial neural network (PNN) based surrogate model is used to analyze the stochastic power output. A sensitivity analysis is conducted and highly influenced parameters to the electric power output are identified. The accuracy and adaptability of the PNN model are established by comparing the results with Monte Carlo simulation (MCS). Further, the stochastic analyses of power output are performed for various degrees of randomness and wind velocities. The obtained results showed that the influence of the electromechanical coefficient on power output is more compared to other parameters.

Stochastic buckling response of variable fiber spacing composite plate under thermal environment

“…Thus, this investigation is further extended to estimate the contribution of independent composite properties to the critical thermal buckling using variance-based global sensitivity analysis. 46,47 Table 10 represents the sensitivity of each input property to the thermal buckling response for three boundary conditions. It can be noted from Table 10 that the thermal expansion coefficient of the matrix ( α m ) is the most sensitive property, followed by the elastic modulus of the matrix ( E m ), fiber ( E f ), and Poisson’s ratio of the matrix ( υ m ), regardless to the boundary conditions and lamination sequences.…”

Uncertainty quantification in free vibration and aeroelastic response of variable angle tow laminated composite plate

“…More details of the variance-based global sensitivity analysis can be found in ref. 55,56 Reliability analysis…”

Uncertain buckling characteristics of thermally loaded and internally defected variable fiber spacing composite laminates

“…To get over MCS's limitations, many researchers use neural network algorithms to perform various uncertainty analyses. [53][54][55][56][57][58][59]…”