Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493373
Fabian Stuckmann, Pasha A. Fistanto, G. P. Vayá
This paper presents an open-source tool, called PATARA, for post-silicon validation of Application-Specific Instruction set Processors (ASIP). PATARA is based on the REVERSI approach, which can significantly speed up the validation time of already fabricated chip designs. It enables the generation of automatic, randomized test-benches, which are then validated by the processor itself. PARATA extends the REVERSI approach by supporting subword parallalism (i.e., microSIMD hardware mechanism), multiple issue slots, and conditional execution. The configuration of PATARA is provided by XML files, which describe the instruction set architecture (ISA) of the target custom processor. Moreover, the modular structure of the PATARA tool enables the insertion of future features. The potential of this tool is shown in a case study, where part of the instruction set architecture of a VLIW-SIMD processor, called KAVUAKA, is verified. The results show that the use of the PATARA tool can achieve a higher code coverage than the previously used hand-written testbenches. It is worth mentioning that a high code coverage in pre-silicon verification generally implies a high coverage in post-silicon validation.
{"title":"PATARA: A REVERSI-Based Open-Source Tool for Post-Silicon Validation of Processor Cores","authors":"Fabian Stuckmann, Pasha A. Fistanto, G. P. Vayá","doi":"10.1109/MOCAST52088.2021.9493373","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493373","url":null,"abstract":"This paper presents an open-source tool, called PATARA, for post-silicon validation of Application-Specific Instruction set Processors (ASIP). PATARA is based on the REVERSI approach, which can significantly speed up the validation time of already fabricated chip designs. It enables the generation of automatic, randomized test-benches, which are then validated by the processor itself. PARATA extends the REVERSI approach by supporting subword parallalism (i.e., microSIMD hardware mechanism), multiple issue slots, and conditional execution. The configuration of PATARA is provided by XML files, which describe the instruction set architecture (ISA) of the target custom processor. Moreover, the modular structure of the PATARA tool enables the insertion of future features. The potential of this tool is shown in a case study, where part of the instruction set architecture of a VLIW-SIMD processor, called KAVUAKA, is verified. The results show that the use of the PATARA tool can achieve a higher code coverage than the previously used hand-written testbenches. It is worth mentioning that a high code coverage in pre-silicon verification generally implies a high coverage in post-silicon validation.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121533649","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493385
Dimitris Nikitas, Konstantinos Papafotis, P. Sotiriadis
In this work we propose a near-optimal Bang-Bang controller for the longitudinal motion of a quadrotor UAV which minimizes the active flight time. The proposed controller is closed loop, has simple cascaded structure and can be used in online navigation since it has minimum computational requirements compared to other approaches in the literature. It is evaluated via numerical simulation for a standard quadrotor nonlinear model.
{"title":"Time-Near-Optimal Longitudinal Control for Quadrotor UAVs","authors":"Dimitris Nikitas, Konstantinos Papafotis, P. Sotiriadis","doi":"10.1109/MOCAST52088.2021.9493385","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493385","url":null,"abstract":"In this work we propose a near-optimal Bang-Bang controller for the longitudinal motion of a quadrotor UAV which minimizes the active flight time. The proposed controller is closed loop, has simple cascaded structure and can be used in online navigation since it has minimum computational requirements compared to other approaches in the literature. It is evaluated via numerical simulation for a standard quadrotor nonlinear model.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124146306","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493400
J. Vourvoulakis, Leonardo Bilalis
Pulse oximetry is a popular non-invasive method for monitoring the oxygen saturation levels (SpO2) in blood as well as the heart rate (HR) of a patient. The photoplethysmographic signal (PPG) is used to estimate SpO2 and HR. It indicates the light absorption of oxygenated and deoxygenated hemoglobin at a certain wavelength. Red and IR wavelengths are often used to extract PPG signals. Subsequently, various algorithms can be applied to the PPG signals in order to obtain SpO2 and HR. In this paper, we propose a pulse oximetry system in which a lightweight algorithm is applied for HR and SpO2 estimation. Our study was based on PPG signals derived from the MAX30102 sensor. PIC18F46Q43 microcontroller unit (MCU) was selected as the system processor, which was responsible for the sensor readouts and for the implementation of the algorithm. Communication with each external device was accomplished by using Direct Memory Access (DMA) transfers. Furthermore, the required functionality was deployed by adopting a task pipeline firmware scheme. In that scheme, operations were completed in parallel. This technique accelerated the execution and maximized the time in which the MCU can be put in low power mode. Interconnection between our system and a personal computer was also realized by using an external USB-to-serial module. Associated Octave scripts for receiving, analyzing and processing of PPG signal data were also developed.
{"title":"Real-time pulse oximetry extraction using a lightweight algorithm and a task pipeline scheme","authors":"J. Vourvoulakis, Leonardo Bilalis","doi":"10.1109/MOCAST52088.2021.9493400","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493400","url":null,"abstract":"Pulse oximetry is a popular non-invasive method for monitoring the oxygen saturation levels (SpO2) in blood as well as the heart rate (HR) of a patient. The photoplethysmographic signal (PPG) is used to estimate SpO2 and HR. It indicates the light absorption of oxygenated and deoxygenated hemoglobin at a certain wavelength. Red and IR wavelengths are often used to extract PPG signals. Subsequently, various algorithms can be applied to the PPG signals in order to obtain SpO2 and HR. In this paper, we propose a pulse oximetry system in which a lightweight algorithm is applied for HR and SpO2 estimation. Our study was based on PPG signals derived from the MAX30102 sensor. PIC18F46Q43 microcontroller unit (MCU) was selected as the system processor, which was responsible for the sensor readouts and for the implementation of the algorithm. Communication with each external device was accomplished by using Direct Memory Access (DMA) transfers. Furthermore, the required functionality was deployed by adopting a task pipeline firmware scheme. In that scheme, operations were completed in parallel. This technique accelerated the execution and maximized the time in which the MCU can be put in low power mode. Interconnection between our system and a personal computer was also realized by using an external USB-to-serial module. Associated Octave scripts for receiving, analyzing and processing of PPG signal data were also developed.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127494968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493394
G. Orfanoudakis, E. Koutroulis, G. Foteinopoulos
Conventional DC/AC inverters used for the integration of photovoltaic (PV) energy sources to the utility grid are known to give rise to ground leakage currents, which can damage the PV cells and cause safety hazards. A modern solution to the leakage current problem is offered by transformerless PV inverter topologies. This paper focuses on the class of decoupling transformerless topologies and provides a complete explanation of their leakage current suppression mechanism, by considering the commonly neglected role of the inverter diodes. It is demonstrated that, apart from additional semiconductor switches, the diodes are essential for achieving leakage current suppression, while the leakage current that flows through them radically affects the waveform of the inverter common-mode voltage. The presented analysis is supported by simulation results in MATLAB-Simulink.
{"title":"The role of diodes in the leakage current suppression mechanism of decoupling transformerless PV inverter topologies","authors":"G. Orfanoudakis, E. Koutroulis, G. Foteinopoulos","doi":"10.1109/MOCAST52088.2021.9493394","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493394","url":null,"abstract":"Conventional DC/AC inverters used for the integration of photovoltaic (PV) energy sources to the utility grid are known to give rise to ground leakage currents, which can damage the PV cells and cause safety hazards. A modern solution to the leakage current problem is offered by transformerless PV inverter topologies. This paper focuses on the class of decoupling transformerless topologies and provides a complete explanation of their leakage current suppression mechanism, by considering the commonly neglected role of the inverter diodes. It is demonstrated that, apart from additional semiconductor switches, the diodes are essential for achieving leakage current suppression, while the leakage current that flows through them radically affects the waveform of the inverter common-mode voltage. The presented analysis is supported by simulation results in MATLAB-Simulink.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124270395","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493415
Theodora Sanida, Dimitris Tsiktsiris, Argyrios Sideris, M. Dasygenis
The economy of many countries worldwide depends on the agricultural sector. In agriculture, plant cultivation is often affected by diseases. The plant diseases can cause partial or complete damage to the leaves and fruit. This leads to significant economic loss in the agricultural industry. At the same time, plant diseases, if not detected in an early stage, have negative consequences on the quantity, quality and production of agricultural crops. Thus, the biggest challenge in agriculture is the accurate and early diagnosis and the prevention of plant diseases. In this study, we modified the structure of a popular classification network, the MobileNet V2 to improve the overall accuracy of it, by applying the transfer learning technique and fine-tuning. Experimental findings have shown that the Modified MobileNet V2 architecture achieves a high level of overall accuracy compared to other similar CNN lightweight architectures used in plant leaf disease detection in the recent literature.
{"title":"A Heterogeneous Lightweight Network for Plant Disease Classification","authors":"Theodora Sanida, Dimitris Tsiktsiris, Argyrios Sideris, M. Dasygenis","doi":"10.1109/MOCAST52088.2021.9493415","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493415","url":null,"abstract":"The economy of many countries worldwide depends on the agricultural sector. In agriculture, plant cultivation is often affected by diseases. The plant diseases can cause partial or complete damage to the leaves and fruit. This leads to significant economic loss in the agricultural industry. At the same time, plant diseases, if not detected in an early stage, have negative consequences on the quantity, quality and production of agricultural crops. Thus, the biggest challenge in agriculture is the accurate and early diagnosis and the prevention of plant diseases. In this study, we modified the structure of a popular classification network, the MobileNet V2 to improve the overall accuracy of it, by applying the transfer learning technique and fine-tuning. Experimental findings have shown that the Modified MobileNet V2 architecture achieves a high level of overall accuracy compared to other similar CNN lightweight architectures used in plant leaf disease detection in the recent literature.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"110 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122124826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493371
Athanasios Girgolas, C. Volos, A. Giakoumis, S. Stavrinides, T. Karakasidis, I. Stouboulos
In this work, for the first time a macroeconomic model has been emulated using a microcontroller-based device, considering as case study the Greek economic crises of period 2010-2019. With the use of well-known tools from nonlinear theory, such as phase portrait, bifurcation diagram, Poincaré map and Lyapunov exponents, interesting phenomena have been revealed. More precisely, the period-doubling route to chaos, crisis phenomena and coexistence of attractors have been presented in detail.
{"title":"Emulating a Chaotic Economic Model By Using A Microcontroller","authors":"Athanasios Girgolas, C. Volos, A. Giakoumis, S. Stavrinides, T. Karakasidis, I. Stouboulos","doi":"10.1109/MOCAST52088.2021.9493371","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493371","url":null,"abstract":"In this work, for the first time a macroeconomic model has been emulated using a microcontroller-based device, considering as case study the Greek economic crises of period 2010-2019. With the use of well-known tools from nonlinear theory, such as phase portrait, bifurcation diagram, Poincaré map and Lyapunov exponents, interesting phenomena have been revealed. More precisely, the period-doubling route to chaos, crisis phenomena and coexistence of attractors have been presented in detail.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116586463","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493364
A. Slavova, V. Ignatov
In this work we present a special class of memristor Cellular Nonlinear/Nanoscale Networks (MCNN) operating on the edge of chaos. We study the dynamics of Gierer-Menhardt system for pattern formation in living cells. We determine three regions in the parameter set: locally active and stable, locally active and unstable and locally passive. Throughout extensive simulations we obtain non-uniform spatial-pattern generation of MCNN model in the edge of chaos domain. Applications in EEG signal generation are shown in order to predict the epileptic seizures.
{"title":"Universal Cellular Computing on the Edge of Chaos","authors":"A. Slavova, V. Ignatov","doi":"10.1109/MOCAST52088.2021.9493364","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493364","url":null,"abstract":"In this work we present a special class of memristor Cellular Nonlinear/Nanoscale Networks (MCNN) operating on the edge of chaos. We study the dynamics of Gierer-Menhardt system for pattern formation in living cells. We determine three regions in the parameter set: locally active and stable, locally active and unstable and locally passive. Throughout extensive simulations we obtain non-uniform spatial-pattern generation of MCNN model in the edge of chaos domain. Applications in EEG signal generation are shown in order to predict the epileptic seizures.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"286 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124562639","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493377
V. Gogolou, Zoi Agorastou, V. Kalenteridis, Konstantinos Kozalakis, I. Kosmadakis, K. Siozios, E. Koutroulis, S. Siskos
In this paper, the challenges of designing a boost converter in standard CMOS technology are discussed. Based on theoretical calculations, which take into consideration the characteristics of the technology used, efficient solutions are proposed to overcome the current-stress limitations as well as the power dissipation issues. Following these guidelines, a test DC-DC converter, designed and fabricated in a standard 0.18um CMOS process is presented, along with experimental results.
{"title":"Design considerations for a DC-DC Boost Converter in standard CMOS technology","authors":"V. Gogolou, Zoi Agorastou, V. Kalenteridis, Konstantinos Kozalakis, I. Kosmadakis, K. Siozios, E. Koutroulis, S. Siskos","doi":"10.1109/MOCAST52088.2021.9493377","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493377","url":null,"abstract":"In this paper, the challenges of designing a boost converter in standard CMOS technology are discussed. Based on theoretical calculations, which take into consideration the characteristics of the technology used, efficient solutions are proposed to overcome the current-stress limitations as well as the power dissipation issues. Following these guidelines, a test DC-DC converter, designed and fabricated in a standard 0.18um CMOS process is presented, along with experimental results.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121161458","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493342
Iason Kalaitzakis, Michail Dakanalis, F. Kanellos
The increasing popularity of Plug-in Electric Vehicles (PEVs) will impose a significant strain on the electricity networks. However, Vehicle-to-grid (V2G) operation can provide ancillary services, improving the overall stability and reliability of the grid. In this paper, an optimal power management system with frequency support capability is proposed. Particle Swarm Optimization (PSO) algorithm is used for the calculation of the optimal power exchange between PEV’s battery and the grid throughout the day. Additionally, the ability of the PEV to provide frequency support to the grid is examined. Using a frequency-PEV load droop control approach, the deviation of PEV’s power from its optimal set-point is determined. Once the need for frequency support ends, the aforementioned algorithm reschedules the new optimal operation of the PEV, satisfying at the same time all predetermined goals and constraints.
{"title":"Optimal Power Management for Residential PEV Chargers with Frequency Support Capability","authors":"Iason Kalaitzakis, Michail Dakanalis, F. Kanellos","doi":"10.1109/MOCAST52088.2021.9493342","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493342","url":null,"abstract":"The increasing popularity of Plug-in Electric Vehicles (PEVs) will impose a significant strain on the electricity networks. However, Vehicle-to-grid (V2G) operation can provide ancillary services, improving the overall stability and reliability of the grid. In this paper, an optimal power management system with frequency support capability is proposed. Particle Swarm Optimization (PSO) algorithm is used for the calculation of the optimal power exchange between PEV’s battery and the grid throughout the day. Additionally, the ability of the PEV to provide frequency support to the grid is examined. Using a frequency-PEV load droop control approach, the deviation of PEV’s power from its optimal set-point is determined. Once the need for frequency support ends, the aforementioned algorithm reschedules the new optimal operation of the PEV, satisfying at the same time all predetermined goals and constraints.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128526997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-05DOI: 10.1109/MOCAST52088.2021.9493355
D. Kampelopoulos, Georgios-Panagiotis Kousiopoulos, N. Karagiorgos, V. Konstantakos, S. Goudos, S. Nikolaidis
In this work, a machine learning approach is proposed for the problem of leak detection in noisy industrial pipelines. The traditional multi-class or binary classification approaches depend on the fact that real training data are required. However, in real pipeline scenarios the data generation for the leak class relies on measurements of artificially generated leaks which are different in nature than actual ones. Also, some pipelines are not equipped with the components to generate these leaks and in some cases, it is difficult to acquire a large and balanced leak dataset. Thus, in this paper, a set of one class classification models are applied that do not require training with real leak data. In this study's case, four one class classification models are trained on a single class representing the pipeline's normal operating noise. Seven time and frequency domain features are extracted from the raw acoustic data acquired by a set of accelerometers. The trained models are then tested on a new dataset containing leak and noise measurements. This dataset is used to evaluate each model's ability to detect leaks as well as the effect that some introduced parameters have on their performance. Overall, high levels of accuracy are exhibited, and all models are able to distinguish between noise and leak data.
{"title":"Applying One Class Classification for Leak Detection in Noisy Industrial Pipelines","authors":"D. Kampelopoulos, Georgios-Panagiotis Kousiopoulos, N. Karagiorgos, V. Konstantakos, S. Goudos, S. Nikolaidis","doi":"10.1109/MOCAST52088.2021.9493355","DOIUrl":"https://doi.org/10.1109/MOCAST52088.2021.9493355","url":null,"abstract":"In this work, a machine learning approach is proposed for the problem of leak detection in noisy industrial pipelines. The traditional multi-class or binary classification approaches depend on the fact that real training data are required. However, in real pipeline scenarios the data generation for the leak class relies on measurements of artificially generated leaks which are different in nature than actual ones. Also, some pipelines are not equipped with the components to generate these leaks and in some cases, it is difficult to acquire a large and balanced leak dataset. Thus, in this paper, a set of one class classification models are applied that do not require training with real leak data. In this study's case, four one class classification models are trained on a single class representing the pipeline's normal operating noise. Seven time and frequency domain features are extracted from the raw acoustic data acquired by a set of accelerometers. The trained models are then tested on a new dataset containing leak and noise measurements. This dataset is used to evaluate each model's ability to detect leaks as well as the effect that some introduced parameters have on their performance. Overall, high levels of accuracy are exhibited, and all models are able to distinguish between noise and leak data.","PeriodicalId":146990,"journal":{"name":"2021 10th International Conference on Modern Circuits and Systems Technologies (MOCAST)","volume":"165 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127352437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：