Deepesh Upadrashta
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Deepesh Upadrashta has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Yaowen Yang
School of Civil and Environmental Engineering, Nanyang Technological University, Singapore 639798, Singapore
7 shared publications Xiangyang Li
4 shared publications- Kaiping Yu
School of Astronautics, Harbin Institute of Technology, Harbin, China
3 shared publications - Jinda Jia
School of Mechatronics Engineering, Harbin Institute of Technology, Harbin 150001, China
1 shared publication Kaiping Yu
1 shared publicationM. Krishnasamy
Microelectronics and VLSI Design Group, Department of Electronics and Communication Engineering, National Institute of Technology, Silchar, Assam 788010, India, and also with the School of Electronics and Electrical Technology, Kalasalingam Academy of Research and Education, Krishnankovil 626126, India
1 shared publication
Latest Publications
- Open Access
- 24 December 2025
- Published by MDPI AG in Information
- Vol. 17 (1) , 16
AI/ML Based Anomaly Detection and Fault Diagnosis of Turbocharged Marine Diesel Engines: Experimental Study on Engine of an Operational Vessel
Turbocharged diesel engines are widely used for the propulsion and as the generators for powering auxiliary systems in marine applications. Many works were published on the development of diagnosis tools for the engines using data from simulation models or from experiments on a sophisticated engine test bench. However, the simulation data varies a lot with actual operational data, and the available sensor data on the actual vessel is much less compared to the data from test benches. Therefore, it is necessary to develop anomaly prediction and fault diagnosis models from limited data available from the engines. In this paper, an artificial intelligence (AI)-based anomaly detection model and machine learning (ML)-based fault diagnosis model were developed using the actual data acquired from a diesel engine of a cargo vessel. Unlike the previous works, the study uses operational, thermodynamic, and vibration data for the anomaly detection and fault diagnosis. The paper provides the overall architecture of the proposed predictive maintenance system including details on the sensorization of assets, data acquisition, edge computation, and AI model for anomaly prediction and ML algorithm for fault diagnosis. Faults with varying severity levels were induced in the subcomponents of the engine to validate the accuracy of the anomaly detection and fault diagnosis models. The unsupervised stacked autoencoder AI model predicts the engine anomalies with 87.6% accuracy. The balanced accuracy of supervised fault diagnosis model using Support Vector Machine algorithm is 99.7%. The proposed models are vital in marching towards sustainable shipping and have potential to deploy across various applications.
- 24 December 2025
- Published by MDPI AG in Information
- Vol. 17 (1) , 16
Open AccessAI/ML Based Anomaly Detection and Fault Diagnosis of Turbocharged Marine Diesel Engines: Experimental Study on Engine of an Operational Vessel
Turbocharged diesel engines are widely used for the propulsion and as the generators for powering auxiliary systems in marine applications. Many works were published on the development of diagnosis tools for the engines using data from simulation models or from experiments on a sophisticated engine test bench. However, the simulation data varies a lot with actual operational data, and the available sensor data on the actual vessel is much less compared to the data from test benches. Therefore, it is necessary to develop anomaly prediction and fault diagnosis models from limited data available from the engines. In this paper, an artificial intelligence (AI)-based anomaly detection model and machine learning (ML)-based fault diagnosis model were developed using the actual data acquired from a diesel engine of a cargo vessel. Unlike the previous works, the study uses operational, thermodynamic, and vibration data for the anomaly detection and fault diagnosis. The paper provides the overall architecture of the proposed predictive maintenance system including details on the sensorization of assets, data acquisition, edge computation, and AI model for anomaly prediction and ML algorithm for fault diagnosis. Faults with varying severity levels were induced in the subcomponents of the engine to validate the accuracy of the anomaly detection and fault diagnosis models. The unsupervised stacked autoencoder AI model predicts the engine anomalies with 87.6% accuracy. The balanced accuracy of supervised fault diagnosis model using Support Vector Machine algorithm is 99.7%. The proposed models are vital in marching towards sustainable shipping and have potential to deploy across various applications.
- 1 June 2019
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 124, 613-631
Analytical modeling and validation of multi-mode piezoelectric energy harvester
- 1 June 2019
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 124, 613-631
Analytical modeling and validation of multi-mode piezoelectric energy harvester
- 28 March 2019
- Published by IOP Publishing in Journal of Physics D: Applied Physics
- Vol. 52 (23) , 235501
Comparative study of core materials and multi-degree-of-freedom sandwich piezoelectric energy harvester with inner cantilevered beams
Conventional energy harvesters using a single metal substrate can hardly collect sufficient energies from wideband and low frequency vibration sources. A sandwich structure composed of two thin face sheets and a soft-core material can be an alternative substrate to replace a single metal layer to drop the resonant frequency of the harvester accompanied by the increment of the output voltage. The resonant frequency of the sandwich piezoelectric energy harvester (SPEH) can be adjusted more flexibly by selecting the appropriate core material and varying the geometric dimensions of the sandwich substrate. In this paper, a comparative study of different core materials for SPEH is firstly investigated. Prototypes of the SPEHs with different core materials (natural rubber, silicon rubber, polycarbonate plastic) and conventional harvesters with the same geometrical dimension are fabricated and tested. The experiment displays that the plastic sandwich energy harvester is the best of all in terms of voltage output. To extend the application of SPEH to harvester broadband vibration energies, this paper presents a novel multi-degree-of-freedom sandwich energy harvester with inner cantilevered beams to harvester energy from wideband, low frequency and low amplitude vibration sources. The proposed energy harvester comprises of a partitioned main sandwich substrate with a tip mass and two patches of piezoelectric layer bonded on it. Multiple branches with tip masses are extended from the inner substrate. The finite element model of the proposed sandwich harvester is presented to achieve close resonant frequencies in the low frequency range of interest. Then a multi-degree-of-freedom sandwich harvester with inner cantilevered beams is designed, fabricated and tested under. The novel harvester proposed in this paper has high design flexibility to work in various vibration environments.
- 28 March 2019
- Published by IOP Publishing in Journal of Physics D: Applied Physics
- Vol. 52 (23) , 235501
Comparative study of core materials and multi-degree-of-freedom sandwich piezoelectric energy harvester with inner cantilevered beams
Conventional energy harvesters using a single metal substrate can hardly collect sufficient energies from wideband and low frequency vibration sources. A sandwich structure composed of two thin face sheets and a soft-core material can be an alternative substrate to replace a single metal layer to drop the resonant frequency of the harvester accompanied by the increment of the output voltage. The resonant frequency of the sandwich piezoelectric energy harvester (SPEH) can be adjusted more flexibly by selecting the appropriate core material and varying the geometric dimensions of the sandwich substrate. In this paper, a comparative study of different core materials for SPEH is firstly investigated. Prototypes of the SPEHs with different core materials (natural rubber, silicon rubber, polycarbonate plastic) and conventional harvesters with the same geometrical dimension are fabricated and tested. The experiment displays that the plastic sandwich energy harvester is the best of all in terms of voltage output. To extend the application of SPEH to harvester broadband vibration energies, this paper presents a novel multi-degree-of-freedom sandwich energy harvester with inner cantilevered beams to harvester energy from wideband, low frequency and low amplitude vibration sources. The proposed energy harvester comprises of a partitioned main sandwich substrate with a tip mass and two patches of piezoelectric layer bonded on it. Multiple branches with tip masses are extended from the inner substrate. The finite element model of the proposed sandwich harvester is presented to achieve close resonant frequencies in the low frequency range of interest. Then a multi-degree-of-freedom sandwich harvester with inner cantilevered beams is designed, fabricated and tested under. The novel harvester proposed in this paper has high design flexibility to work in various vibration environments.
- 1 February 2019
- Published by IOP Publishing in Smart Materials and Structures
- Vol. 28 (3) , 035010
Multi-branch sandwich piezoelectric energy harvester: mathematical modeling and validation
Energy harvesting from ambient environment has attracted intensive attention over the years for powering low-power autonomous electronic devices. A conventional piezoelectric energy harvester (CPEH) can hardly meet the requirements of effective energy harvesting from wideband, low frequency and low amplitude vibration sources due to its simple substrate and single resonant peak response. Using sandwich structure comprised of a soft-core layer and two thin skins can drastically reduce the natural frequency and increase voltage output of the harvester. The stiffness of sandwich harvester and consequently its resonant frequency can be tuned more flexibly than the CPEH by choosing different materials and adjusting the geometric dimensions of the core and skins. In this paper, a novel multi-branch sandwich piezoelectric energy harvester (MSPEH) is proposed to harvest energy from wideband, low frequency and low amplitude vibration sources. The proposed harvester comprises of a main sandwich beam as substrate with a patch of piezoelectric layer bonded over it. Multiple inner single branches with tip masses are connected to the main substrate to generate multiple resonant peaks and tune the range of the interested frequency. Firstly, a mathematical model of the proposed harvester is presented, and the electromechanical coupling equations are obtained. Subsequently, a prototype of MSPEH with two inner single branches is fabricated and tested under harmonic base excitation to study its performance. When tested at a low harmonic excitation of 0.02 g, the harvester generates 2.48 V, 6.21 V and 1.55 V at three resonant frequencies of 18.18 Hz, 24.74 Hz and 28.12 Hz, respectively. Finally, the accuracy of derived mathematical model is verified by comparing with the predictions from finite element simulation and experimental results. The novel MSPEH offers excellent design flexibility to adjust resonant frequencies by choosing inner single branches and tip masses and thus it has potential to generate sufficient power output from wideband, low frequency and low amplitude vibration sources.
- 1 February 2019
- Published by IOP Publishing in Smart Materials and Structures
- Vol. 28 (3) , 035010
Multi-branch sandwich piezoelectric energy harvester: mathematical modeling and validation
Energy harvesting from ambient environment has attracted intensive attention over the years for powering low-power autonomous electronic devices. A conventional piezoelectric energy harvester (CPEH) can hardly meet the requirements of effective energy harvesting from wideband, low frequency and low amplitude vibration sources due to its simple substrate and single resonant peak response. Using sandwich structure comprised of a soft-core layer and two thin skins can drastically reduce the natural frequency and increase voltage output of the harvester. The stiffness of sandwich harvester and consequently its resonant frequency can be tuned more flexibly than the CPEH by choosing different materials and adjusting the geometric dimensions of the core and skins. In this paper, a novel multi-branch sandwich piezoelectric energy harvester (MSPEH) is proposed to harvest energy from wideband, low frequency and low amplitude vibration sources. The proposed harvester comprises of a main sandwich beam as substrate with a patch of piezoelectric layer bonded over it. Multiple inner single branches with tip masses are connected to the main substrate to generate multiple resonant peaks and tune the range of the interested frequency. Firstly, a mathematical model of the proposed harvester is presented, and the electromechanical coupling equations are obtained. Subsequently, a prototype of MSPEH with two inner single branches is fabricated and tested under harmonic base excitation to study its performance. When tested at a low harmonic excitation of 0.02 g, the harvester generates 2.48 V, 6.21 V and 1.55 V at three resonant frequencies of 18.18 Hz, 24.74 Hz and 28.12 Hz, respectively. Finally, the accuracy of derived mathematical model is verified by comparing with the predictions from finite element simulation and experimental results. The novel MSPEH offers excellent design flexibility to adjust resonant frequencies by choosing inner single branches and tip masses and thus it has potential to generate sufficient power output from wideband, low frequency and low amplitude vibration sources.
- Open Access
- 17 December 2018
- Published by MDPI AG in Micromachines
- Vol. 9 (12) , 667
Modeling and Analysis of Upright Piezoelectric Energy Harvester under Aerodynamic Vortex-induced Vibration
This paper presents an upright piezoelectric energy harvester (UPEH) with cylinder extension along its longitudinal direction. The UPEH can generate energy from low-speed wind by bending deformation produced by vortex-induced vibrations (VIVs). The UPEH has the advantages of less working space and ease of setting up an array over conventional vortex-induced vibration harvesters. The nonlinear distributed modeling method is established based on Euler–Bernoulli beam theory and aerodynamic vortex-induced force of the cylinder is obtained by the van der Pol wake oscillator theory. The fluid–solid–electricity governing coupled equations are derived using Lagrange’s equation and solved through Galerkin discretization. The effect of cylinder gravity on the dynamic characteristics of the UPEH is also considered using the energy method. The influences of substrate dimension, piezoelectric dimension, the mass of cylinder extension, and electrical load resistance on the output performance of harvester are studied using the theoretical model. Experiments were carried out and the results were in good agreement with the numerical results. The results showed that a UPEH configuration achieves the maximum power of 635.04 μW at optimum resistance of 250 kΩ when tested at a wind speed of 4.20 m/s. The theoretical results show that the UPEH can get better energy harvesting output performance with a lighter tip mass of cylinder, and thicker and shorter substrate in its synchronization working region. This work will provide the theoretical guidance for studying the array of multiple upright energy harvesters.
- 17 December 2018
- Published by MDPI AG in Micromachines
- Vol. 9 (12) , 667
Open AccessModeling and Analysis of Upright Piezoelectric Energy Harvester under Aerodynamic Vortex-induced Vibration
This paper presents an upright piezoelectric energy harvester (UPEH) with cylinder extension along its longitudinal direction. The UPEH can generate energy from low-speed wind by bending deformation produced by vortex-induced vibrations (VIVs). The UPEH has the advantages of less working space and ease of setting up an array over conventional vortex-induced vibration harvesters. The nonlinear distributed modeling method is established based on Euler–Bernoulli beam theory and aerodynamic vortex-induced force of the cylinder is obtained by the van der Pol wake oscillator theory. The fluid–solid–electricity governing coupled equations are derived using Lagrange’s equation and solved through Galerkin discretization. The effect of cylinder gravity on the dynamic characteristics of the UPEH is also considered using the energy method. The influences of substrate dimension, piezoelectric dimension, the mass of cylinder extension, and electrical load resistance on the output performance of harvester are studied using the theoretical model. Experiments were carried out and the results were in good agreement with the numerical results. The results showed that a UPEH configuration achieves the maximum power of 635.04 μW at optimum resistance of 250 kΩ when tested at a wind speed of 4.20 m/s. The theoretical results show that the UPEH can get better energy harvesting output performance with a lighter tip mass of cylinder, and thicker and shorter substrate in its synchronization working region. This work will provide the theoretical guidance for studying the array of multiple upright energy harvesters.
- 1 November 2018
- Published by Elsevier in Energy Conversion and Management
- Vol. 176, 69-85
Sandwich piezoelectric energy harvester: Analytical modeling and experimental validation
- 1 November 2018
- Published by Elsevier in Energy Conversion and Management
- Vol. 176, 69-85
Sandwich piezoelectric energy harvester: Analytical modeling and experimental validation
- 26 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Microelectromechanical Systems
- Vol. 27 (6) , 1160-1170
Distributed Parameter Modelling of Cutout 2-DOF Cantilevered Piezo-Magneto-Elastic Energy Harvester
Energy harvesters play the vital role in replacing/recharging the electrochemical batteries in low-power electronic devices. Non-linear piezoelectric energy harvesters generate electrical power from broadband frequency spectrum using ambient vibrational sources. Non-linear cutout 2- degree of freedom (DOF) piezo-magneto-elastic (PME) energy harvester has been proposed to enhance the bandwidth where two close resonant peaks are achieved using magnetic interaction. In this paper, the novel theoretical distributed parametric model of cutout 2-DOF energy harvester is presented which can predict the dynamic behavior of the linear (without magnetic force) and non-linear monostable configurations accurately. The closed-form analytical solutions are derived for the harvested electrical responses and displacements using forcing function, forward and backward coupling factors through Extended Hamilton principle, Galerkin approach, and Euler-Lagrange equations. Specifically, the static displacements and tuned resonant frequencies are obtained as a function of the initial distance between the two magnets in attractive and repulsive configurations. The theoretical findings are validated with COMSOL simulation results and experimental data. The comparison among theoretical, finite-element, and experimental results show that the developed model of cutout 2-DOF piezo-magneto-elastic energy harvester can accurately predict the response for linear and non-linear monostable configurations. [2018-0068]
- 26 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Microelectromechanical Systems
- Vol. 27 (6) , 1160-1170
Distributed Parameter Modelling of Cutout 2-DOF Cantilevered Piezo-Magneto-Elastic Energy Harvester
Energy harvesters play the vital role in replacing/recharging the electrochemical batteries in low-power electronic devices. Non-linear piezoelectric energy harvesters generate electrical power from broadband frequency spectrum using ambient vibrational sources. Non-linear cutout 2- degree of freedom (DOF) piezo-magneto-elastic (PME) energy harvester has been proposed to enhance the bandwidth where two close resonant peaks are achieved using magnetic interaction. In this paper, the novel theoretical distributed parametric model of cutout 2-DOF energy harvester is presented which can predict the dynamic behavior of the linear (without magnetic force) and non-linear monostable configurations accurately. The closed-form analytical solutions are derived for the harvested electrical responses and displacements using forcing function, forward and backward coupling factors through Extended Hamilton principle, Galerkin approach, and Euler-Lagrange equations. Specifically, the static displacements and tuned resonant frequencies are obtained as a function of the initial distance between the two magnets in attractive and repulsive configurations. The theoretical findings are validated with COMSOL simulation results and experimental data. The comparison among theoretical, finite-element, and experimental results show that the developed model of cutout 2-DOF piezo-magneto-elastic energy harvester can accurately predict the response for linear and non-linear monostable configurations. [2018-0068]
- 1 September 2018
- Published by American Society of Civil Engineers (ASCE) in Journal of Aerospace Engineering
- Vol. 31 (5)
Trident-Shaped Multimodal Piezoelectric Energy Harvester
Energy harvesting from ambient vibrations using piezoelectric materials has garnered much research focus over the years due to its immense potential to replace/recharge batteries in wireless sensor nodes and low-power-consuming autonomous electronic devices. Practical implementation of conventional linear piezoelectric energy harvesters is not an effective solution because they suffer from narrow operational bandwidth due to their single resonant peak response in frequency spectrum. Several techniques, such as oscillator arrays, passive/active resonant tuning, and nonlinear methods, have been proposed to broaden the bandwidth of harvesters. This paper proposes a novel trident (three-pronged spear) shaped multimodal (three degrees of freedom) piezoelectric energy harvester to harness electrical energy from wideband, low-frequency, and low-amplitude ambient vibrations. A single patch of piezoelectric material is used for power generation from multiple modes of the structure. The harvester structure consists of a primary cantilever beam with a patch of macrofiber composite bonded on it and three branched beams with tip masses attached to the free end of the primary beam. The first three bending modes of the harvester are used for power generation. The proposed harvester can be designed to operate in the frequency range of the target vibration source. A parametric study varying the geometric parameters of the harvester is conducted using ANSYS finite-element analysis software to obtain the three resonant peaks in the 15–20 Hz bandwidth. The prototype of multimodal harvester is fabricated in accordance with the parametric study and tested under harmonic and random excitations. When tested at harmonic excitation, the harvester generates 3738, 47, and 123 μW at the three resonant peaks. When tested under random excitation, the harvester accumulates sufficient energy in a capacitor to support a low-power standalone wireless sensing unit.- 1 September 2018
- Published by American Society of Civil Engineers (ASCE) in Journal of Aerospace Engineering
- Vol. 31 (5)
Trident-Shaped Multimodal Piezoelectric Energy Harvester
Energy harvesting from ambient vibrations using piezoelectric materials has garnered much research focus over the years due to its immense potential to replace/recharge batteries in wireless sensor nodes and low-power-consuming autonomous electronic devices. Practical implementation of conventional linear piezoelectric energy harvesters is not an effective solution because they suffer from narrow operational bandwidth due to their single resonant peak response in frequency spectrum. Several techniques, such as oscillator arrays, passive/active resonant tuning, and nonlinear methods, have been proposed to broaden the bandwidth of harvesters. This paper proposes a novel trident (three-pronged spear) shaped multimodal (three degrees of freedom) piezoelectric energy harvester to harness electrical energy from wideband, low-frequency, and low-amplitude ambient vibrations. A single patch of piezoelectric material is used for power generation from multiple modes of the structure. The harvester structure consists of a primary cantilever beam with a patch of macrofiber composite bonded on it and three branched beams with tip masses attached to the free end of the primary beam. The first three bending modes of the harvester are used for power generation. The proposed harvester can be designed to operate in the frequency range of the target vibration source. A parametric study varying the geometric parameters of the harvester is conducted using ANSYS finite-element analysis software to obtain the three resonant peaks in the 15–20 Hz bandwidth. The prototype of multimodal harvester is fabricated in accordance with the parametric study and tested under harmonic and random excitations. When tested at harmonic excitation, the harvester generates 3738, 47, and 123 μW at the three resonant peaks. When tested under random excitation, the harvester accumulates sufficient energy in a capacitor to support a low-power standalone wireless sensing unit.