Litcius/Paper detail

Application of artificial intelligence and evolutionary algorithms in simulation-based optimal design of a piezoelectric energy harvester

Shahriar Bagheri, Nan Wu, Shaahin Filizadeh

2020Smart Materials and Structures29 citationsDOI

Abstract

Abstract This paper tackles the problem of finding the optimal design parameters for a piezoelectric energy harvester. A new simulation-based optimization procedure is proposed with the goal of acquiring the optimal geometric and circuit design parameters that leads to higher energy harvesting efficiency and also enhances the obtained electrical power. The basis of the optimization platform is a numerical model of the energy harvesting system operating during electrical transient of charging an external storage capacitor. The model consists of a cantilever beam partially coated with piezoelectric patches, a non-linear interfacing and conditioning circuit, and a storage device. The numerical model simulates a complete energy harvesting scenario from piezoelectric transduction, to power enhancement and conditioning through interfacing circuit and energy storage. Two different case studies are considered for beams under harmonic tip-force, and harmonic base-excitation. Since performing multiple simulations in order to evaluate the objective function is computationally expensive and imposes time and space (memory) complexities, a more efficient Neural Network (NN) model is first trained based on a set of training data obtained from the numerical model. Performance and accuracy of the NN training is studied using available statistical methods. Second, a Genetic Algorithm (GA) optimization performs a block-box optimization procedure, using the trained Neural Network model for objective function evaluation. Finally, a thorough analysis of the optimal design parameters obtained from the optimization process is provided.

Topics & Concepts

InterfacingArtificial neural networkOptimal designEvolutionary algorithmComputer scienceGenetic algorithmEnergy harvestingElectronic engineeringEngineeringControl theory (sociology)AlgorithmEnergy (signal processing)Artificial intelligenceMathematicsMachine learningComputer hardwareStatisticsControl (management)Innovative Energy Harvesting TechnologiesAdvanced Sensor and Energy Harvesting MaterialsEnergy Harvesting in Wireless Networks
Application of artificial intelligence and evolutionary algorithms in simulation-based optimal design of a piezoelectric energy harvester | Litcius