Pub Date : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060441
Sameera, M. Tariq, M. Rihan
The power output of solar photovoltaic (PV) panels varies significantly and depends on the solar flux, the state of the power regulating apparatus, geographical locations, and environmental factors. Identifying and analysing these factors is essential for applying suitable mitigation techniques to nullify the power-reducing effects. Dust deposition reduces efficiency by up to 60%, depending on the type and amount of the dust matter. For every 1-degree Celsius increase in the solar cell temperature, the electrical efficiency drops by 0.22%. Similarly, as sun irradiation increases by 100 W/m2, the solar cell temperature and output power grow by 3.82 °C and 3.14 W, respectively. PV module performance is susceptible to being impacted by direct or nearby (in the radius of 60 meters) lightning strikes. This induces overvoltage transients in PV modules and in their power conditioning circuitry. Therefore, the focus of this paper is on mitigation of these atmospheric effects on solar PV panels.
{"title":"Identification and Mitigation of Atmospheric Effects on Solar PV Panel","authors":"Sameera, M. Tariq, M. Rihan","doi":"10.1109/ICERECT56837.2022.10060441","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060441","url":null,"abstract":"The power output of solar photovoltaic (PV) panels varies significantly and depends on the solar flux, the state of the power regulating apparatus, geographical locations, and environmental factors. Identifying and analysing these factors is essential for applying suitable mitigation techniques to nullify the power-reducing effects. Dust deposition reduces efficiency by up to 60%, depending on the type and amount of the dust matter. For every 1-degree Celsius increase in the solar cell temperature, the electrical efficiency drops by 0.22%. Similarly, as sun irradiation increases by 100 W/m2, the solar cell temperature and output power grow by 3.82 °C and 3.14 W, respectively. PV module performance is susceptible to being impacted by direct or nearby (in the radius of 60 meters) lightning strikes. This induces overvoltage transients in PV modules and in their power conditioning circuitry. Therefore, the focus of this paper is on mitigation of these atmospheric effects on solar PV panels.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121390110","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060526
P. D., A. Karegowda, G. M., Abhishek Hooli, R. Aparna, Prashant Gk
Appropriate selection of features play a crucial role for refining precision of classification systems. The classification accuracy and training speed may be significantly intensified by elimination of superfluous features. The present paper addresses the high dimensional data analysis problem through feature selection approach for refining the classification accuracy of Thyroid Nodules (TNs) as benign and malignant. Thyroid Ultrasound Images (TUS) containing nodules are first de-speckled and further improved using Canny Edge Detection (CED) method. This process is followed by application of segmentation technique Adaptive Regularized Kernel Fuzzy C-means (ARKFCM) where relevant Area of Interest (AOI) is obtained and using AOI, nineteen texture features are mined. Finally, feature subset selection is carried out using five different search methods- Genetic Search (GS), Best First (BF), Linear Forward Selection (LFS), Greedy Step Wise (GSW), and Subset Size Forward Selection (SSFS). Selected features are assessed using ten different classifiers Bayes Net, Naïve Bayes, Logistic, Multilayer Perceptron, Radial Basis Function, Sequential Minimal Optimization, Instance Based K-nearest neighbor, K-star, J-48 and Random Tree. Experimental evaluation revealed, features listed using five search techniques have boosted performance of all considered classifiers in comparison to their performance using original nineteen features.
{"title":"Enhanced Thyroid Nodule Classification Adopting Significant Features Selection","authors":"P. D., A. Karegowda, G. M., Abhishek Hooli, R. Aparna, Prashant Gk","doi":"10.1109/ICERECT56837.2022.10060526","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060526","url":null,"abstract":"Appropriate selection of features play a crucial role for refining precision of classification systems. The classification accuracy and training speed may be significantly intensified by elimination of superfluous features. The present paper addresses the high dimensional data analysis problem through feature selection approach for refining the classification accuracy of Thyroid Nodules (TNs) as benign and malignant. Thyroid Ultrasound Images (TUS) containing nodules are first de-speckled and further improved using Canny Edge Detection (CED) method. This process is followed by application of segmentation technique Adaptive Regularized Kernel Fuzzy C-means (ARKFCM) where relevant Area of Interest (AOI) is obtained and using AOI, nineteen texture features are mined. Finally, feature subset selection is carried out using five different search methods- Genetic Search (GS), Best First (BF), Linear Forward Selection (LFS), Greedy Step Wise (GSW), and Subset Size Forward Selection (SSFS). Selected features are assessed using ten different classifiers Bayes Net, Naïve Bayes, Logistic, Multilayer Perceptron, Radial Basis Function, Sequential Minimal Optimization, Instance Based K-nearest neighbor, K-star, J-48 and Random Tree. Experimental evaluation revealed, features listed using five search techniques have boosted performance of all considered classifiers in comparison to their performance using original nineteen features.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"82 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114526226","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060736
T. Prakash, Jayasri B. S, Prakash. K.R.
The entire world witnessed the covid-19pandemicinthe year 2020. The actual outbreak of this corona virus was first reported in Wuhan, China and later declared to be epidemic by (WHO) World Health Organization. The whole world was under tremendous pressure in monitoring health, managing, and maintaining hospitals and inventing new drugs. Initially, India was very much worried because of the huge population. The pandemic posed a critical challenge for healthcare sectors, since doctors and nursing professionals were among the most severely affected and it's clear that India must adopt new measures to increase healthcare proportional ratio and adoption of new technologies to manage large population groups. Robotics is one area which may largely always support the segment. The proposed research project emphasized on developing robotic devices with robotic vision, sensors-based motion planning, dynamic obstacle detection, and autonomous navigation in a hospital environment and supported the medical and nursing teams in reducing their workload and improving patient health monitoring, also the research explored multi-robot exploration and integration.
{"title":"Smart Health Monitoring System Using Robotics","authors":"T. Prakash, Jayasri B. S, Prakash. K.R.","doi":"10.1109/ICERECT56837.2022.10060736","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060736","url":null,"abstract":"The entire world witnessed the covid-19pandemicinthe year 2020. The actual outbreak of this corona virus was first reported in Wuhan, China and later declared to be epidemic by (WHO) World Health Organization. The whole world was under tremendous pressure in monitoring health, managing, and maintaining hospitals and inventing new drugs. Initially, India was very much worried because of the huge population. The pandemic posed a critical challenge for healthcare sectors, since doctors and nursing professionals were among the most severely affected and it's clear that India must adopt new measures to increase healthcare proportional ratio and adoption of new technologies to manage large population groups. Robotics is one area which may largely always support the segment. The proposed research project emphasized on developing robotic devices with robotic vision, sensors-based motion planning, dynamic obstacle detection, and autonomous navigation in a hospital environment and supported the medical and nursing teams in reducing their workload and improving patient health monitoring, also the research explored multi-robot exploration and integration.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124501784","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060642
Geetha Rani E, Mounika E, Gopala Krisnan C, Tanuep Bellam, B. P., Kanagavalli Rengaraju
Human beings have the most distinctive feature that is human face. We can exchange somebody faces with anybody else's faces that appear realistic because many have another type of algo is based upon deepfake tech. Deepfake videos / photos is revolutionary subdual of AI tech by using someones human face can overwrite of someones face. More generously, with many different methods based on productive pictures. Unwillingly the overuse of smartphone and organizing by multiple internet web using AI manipulated data is reaching quicker in something which can we see in the 20th century, global danger is made up by these products Deepfakes are digital manipulation techniques that use machine learning to produce misleading videos. Identification is most difficult part to find from the original. Previously, CNN networks were used to perform identify the deep fake verification. Due to the increasing popularity of deep fakes identification of real one is more important find ways to detect manipulated videos that are presented as real ones. In this project, we will study different methods that can be used to detect such images as well as videos. This study shows that they can also be done using a convolutional algorithm known as Efficient Net and Inception Net. In this Paper, we compare various versions of Convolutional Inception Net with various versions of convolutional Efficient Net combined with Vision Transformers and different Data files to obtain best possible results in Deepfake detection. To get the highly accurate percentage to identify the video is fake or real by using efficient net and by inception net. tract)
{"title":"Comparative Analysis of Deepfake Video Detection Using Inception Net and Efficient Net","authors":"Geetha Rani E, Mounika E, Gopala Krisnan C, Tanuep Bellam, B. P., Kanagavalli Rengaraju","doi":"10.1109/ICERECT56837.2022.10060642","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060642","url":null,"abstract":"Human beings have the most distinctive feature that is human face. We can exchange somebody faces with anybody else's faces that appear realistic because many have another type of algo is based upon deepfake tech. Deepfake videos / photos is revolutionary subdual of AI tech by using someones human face can overwrite of someones face. More generously, with many different methods based on productive pictures. Unwillingly the overuse of smartphone and organizing by multiple internet web using AI manipulated data is reaching quicker in something which can we see in the 20th century, global danger is made up by these products Deepfakes are digital manipulation techniques that use machine learning to produce misleading videos. Identification is most difficult part to find from the original. Previously, CNN networks were used to perform identify the deep fake verification. Due to the increasing popularity of deep fakes identification of real one is more important find ways to detect manipulated videos that are presented as real ones. In this project, we will study different methods that can be used to detect such images as well as videos. This study shows that they can also be done using a convolutional algorithm known as Efficient Net and Inception Net. In this Paper, we compare various versions of Convolutional Inception Net with various versions of convolutional Efficient Net combined with Vision Transformers and different Data files to obtain best possible results in Deepfake detection. To get the highly accurate percentage to identify the video is fake or real by using efficient net and by inception net. tract)","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126273259","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060295
Ting Gao
With the development of network communication, computer technology and artificial intelligence, it has become an inevitable development trend to promote “intelligent buildings” characterized by intelligence, comfort and safety. In order to solve the shortcomings of the existing research on intelligent building control system, this paper discusses the functional equation of CAN algorithm, artificial intelligence fuzzy controller and intelligent building, and aims at the intelligent building control system based on artificial intelligence and CAN algorithm. Hardware settings and parameter settings are briefly introduced. And the work flow design of the intelligent building control system structure based on artificial intelligence and CAN algorithm is discussed, and finally the application of the intelligent building control system based on artificial intelligence and CAN algorithm is experimentally tested. The correct number of switch control by artificial intelligence and CAN algorithm in the intelligent building control system is relatively high. The correct number of switch control in the interference environment is less than that in the humid environment, and the number of switch operations for the control system is the least in the humid environment. The error rate of 500 and up to 2500 is less than 5%, thus verifying the superiority of the application of intelligent building control system based on artificial intelligence and CAN algorithm.
{"title":"Intelligent Building Control System Based on Artificial Intelligence and CAN Algorithm","authors":"Ting Gao","doi":"10.1109/ICERECT56837.2022.10060295","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060295","url":null,"abstract":"With the development of network communication, computer technology and artificial intelligence, it has become an inevitable development trend to promote “intelligent buildings” characterized by intelligence, comfort and safety. In order to solve the shortcomings of the existing research on intelligent building control system, this paper discusses the functional equation of CAN algorithm, artificial intelligence fuzzy controller and intelligent building, and aims at the intelligent building control system based on artificial intelligence and CAN algorithm. Hardware settings and parameter settings are briefly introduced. And the work flow design of the intelligent building control system structure based on artificial intelligence and CAN algorithm is discussed, and finally the application of the intelligent building control system based on artificial intelligence and CAN algorithm is experimentally tested. The correct number of switch control by artificial intelligence and CAN algorithm in the intelligent building control system is relatively high. The correct number of switch control in the interference environment is less than that in the humid environment, and the number of switch operations for the control system is the least in the humid environment. The error rate of 500 and up to 2500 is less than 5%, thus verifying the superiority of the application of intelligent building control system based on artificial intelligence and CAN algorithm.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126431181","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10059648
Raman Mishra, S. Saranya, Mohd Shafahad
Malaria is an epizootic illness caused by unicellular parasites. In Two thousand eighteen there were an estimated two hundred twenty-eight million cases of malaria worldwide. Conventional method of diagnosis requires experienced technician and careful perusal to classify between healthy and infected blood cell, which consumes a lot of time and is also prone to human error. With the help of ML and DL we can simulate human intelligence and make better predictions. The main aim of the paper is to compare the machine learning algorithms namely KNN, Decision Tree, Logistic regression and Random forest and implementing transfer learning with deep learning models VGG19, modified Resnet50 to improve the accuracy achieved with machine learning models thus proposing the best model for predicting malaria only by observing by blood cell image rather than doing any staining of blood, thus reducing any expert requirement.
{"title":"Analysis of different Machine Learning and Deep Learning Techniques for Malaria Parasite Detection","authors":"Raman Mishra, S. Saranya, Mohd Shafahad","doi":"10.1109/ICERECT56837.2022.10059648","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10059648","url":null,"abstract":"Malaria is an epizootic illness caused by unicellular parasites. In Two thousand eighteen there were an estimated two hundred twenty-eight million cases of malaria worldwide. Conventional method of diagnosis requires experienced technician and careful perusal to classify between healthy and infected blood cell, which consumes a lot of time and is also prone to human error. With the help of ML and DL we can simulate human intelligence and make better predictions. The main aim of the paper is to compare the machine learning algorithms namely KNN, Decision Tree, Logistic regression and Random forest and implementing transfer learning with deep learning models VGG19, modified Resnet50 to improve the accuracy achieved with machine learning models thus proposing the best model for predicting malaria only by observing by blood cell image rather than doing any staining of blood, thus reducing any expert requirement.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"255 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121436147","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060562
Aparna P M, Jayalaxmi G N, V. Baligar
There has been an overwhelming increase in social media users in today's world. This ever-increasing data of the Social Network poses a challenge for Link Prediction analysis. The association between users that is not present but has a possibility of existing in the future can be predicted by Link Prediction techniques. In Social Networks, Link Prediction can be employed to monitor social interactions & anomalies, suggest friends to the users and also to analyze the influence or detect communities. Link Prediction helps in retaining the users for longer duration and hence there is a boost in the engagement rate. The more accurate the link prediction is the higher the engagement rate of the applications. Social Networks like Facebook, E-business organisations Zomato and Amazon employ Link Prediction in various forms to boost their revenue and user-experience. There are various algorithms that help in calculation of the possibility of link between entities. The algorithm selection will be based on the specific use case requirement of the applications. The authors of this paper discuss Jaccard Coefficient and Resource Allocation Proximity-based algorithms for Link Prediction. The comparative study is conducted for each of the algorithms and it is observed that the combination of both the algorithms yields a better result than either of them.
{"title":"Link Prediction in Social Networks Using Proximity-Based Algorithms","authors":"Aparna P M, Jayalaxmi G N, V. Baligar","doi":"10.1109/ICERECT56837.2022.10060562","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060562","url":null,"abstract":"There has been an overwhelming increase in social media users in today's world. This ever-increasing data of the Social Network poses a challenge for Link Prediction analysis. The association between users that is not present but has a possibility of existing in the future can be predicted by Link Prediction techniques. In Social Networks, Link Prediction can be employed to monitor social interactions & anomalies, suggest friends to the users and also to analyze the influence or detect communities. Link Prediction helps in retaining the users for longer duration and hence there is a boost in the engagement rate. The more accurate the link prediction is the higher the engagement rate of the applications. Social Networks like Facebook, E-business organisations Zomato and Amazon employ Link Prediction in various forms to boost their revenue and user-experience. There are various algorithms that help in calculation of the possibility of link between entities. The algorithm selection will be based on the specific use case requirement of the applications. The authors of this paper discuss Jaccard Coefficient and Resource Allocation Proximity-based algorithms for Link Prediction. The comparative study is conducted for each of the algorithms and it is observed that the combination of both the algorithms yields a better result than either of them.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131923436","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10059921
Saumya Band, Ruturaj Javeri, V. Kale, Abhishek Morope, Shilpa K. Rudrawar
This paper reports a Military Surveillance System based on IoT. The proposed system is specially designed to fulfill the requirements of soldiers on the battlefield. It employs an ESP8266 controller to work along various sensors to operate it using the application. To ensure that proper surveillance is done on the border, an ESP32 Camera is also employed for Live streaming, face detection & face recognition. Further, the proposed system accurately provides the environmental temperature and conditions. It is a cost-effective, safe, and reliable system for military applications.
{"title":"Military Surveillance System Based on IOT","authors":"Saumya Band, Ruturaj Javeri, V. Kale, Abhishek Morope, Shilpa K. Rudrawar","doi":"10.1109/ICERECT56837.2022.10059921","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10059921","url":null,"abstract":"This paper reports a Military Surveillance System based on IoT. The proposed system is specially designed to fulfill the requirements of soldiers on the battlefield. It employs an ESP8266 controller to work along various sensors to operate it using the application. To ensure that proper surveillance is done on the border, an ESP32 Camera is also employed for Live streaming, face detection & face recognition. Further, the proposed system accurately provides the environmental temperature and conditions. It is a cost-effective, safe, and reliable system for military applications.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132074244","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060356
Y. R, V. S
There is a strong need for energy consumption predictions as it is growing rapidly year by year. These forecasts are beneficial for power production and supply companies and even for the country. Although energy is not the only input that determines the level of production and the degree of economic development of a country, it is highly important for economic growth. It is only by consuming a certain amount of energy that countries can achieve a certain level of economic growth. Hence, it is highly significant to predict energy consumption as it is a growth indicator. Machine learning approaches can forecast the future based on past customer energy consumption as well as various other characteristics. As there are a large number of features that affect the hourly energy consumption, this paper proposes a system that mainly uses the extreme gradient boosting algorithm in the analysis and predictions of energy consumption with feature selection and hyperparameter tuning, achieving the results of hourly energy prediction with a relative error of 7.76% and RMSE of 3.31 kWh.
{"title":"Improved Energy Consumption Prediction using XGBoost with Hyperparameter tuning","authors":"Y. R, V. S","doi":"10.1109/ICERECT56837.2022.10060356","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060356","url":null,"abstract":"There is a strong need for energy consumption predictions as it is growing rapidly year by year. These forecasts are beneficial for power production and supply companies and even for the country. Although energy is not the only input that determines the level of production and the degree of economic development of a country, it is highly important for economic growth. It is only by consuming a certain amount of energy that countries can achieve a certain level of economic growth. Hence, it is highly significant to predict energy consumption as it is a growth indicator. Machine learning approaches can forecast the future based on past customer energy consumption as well as various other characteristics. As there are a large number of features that affect the hourly energy consumption, this paper proposes a system that mainly uses the extreme gradient boosting algorithm in the analysis and predictions of energy consumption with feature selection and hyperparameter tuning, achieving the results of hourly energy prediction with a relative error of 7.76% and RMSE of 3.31 kWh.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134099360","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 : 2022-12-26DOI: 10.1109/ICERECT56837.2022.10060249
S. Anita, E. Elakkia, Y. Sukhi, A. Ahamed, V. Saicharan
The use of hardware with four photo resistors and the open source electrical prototyping platform Arduino is explored in relation to the energy-saving method for the dual axis light (solar) tracker. The defined tolerance for two servo motors is proposed in addition to the required tolerance for the light sensors. In this case, servo motors are turned off to conserve energy if the light intensity changes within a specific range during a specific time period. It is also demonstrated that in this situation, the load on the power supply is almost zero. The task includes the design, development, assembly of all mechanical, electrical, and other components, as well as the development of the control theory guiding all module development and responsible for identifying the scenario of required need.
{"title":"Dual Axis Solar Tracking Based Standalone PV System","authors":"S. Anita, E. Elakkia, Y. Sukhi, A. Ahamed, V. Saicharan","doi":"10.1109/ICERECT56837.2022.10060249","DOIUrl":"https://doi.org/10.1109/ICERECT56837.2022.10060249","url":null,"abstract":"The use of hardware with four photo resistors and the open source electrical prototyping platform Arduino is explored in relation to the energy-saving method for the dual axis light (solar) tracker. The defined tolerance for two servo motors is proposed in addition to the required tolerance for the light sensors. In this case, servo motors are turned off to conserve energy if the light intensity changes within a specific range during a specific time period. It is also demonstrated that in this situation, the load on the power supply is almost zero. The task includes the design, development, assembly of all mechanical, electrical, and other components, as well as the development of the control theory guiding all module development and responsible for identifying the scenario of required need.","PeriodicalId":205485,"journal":{"name":"2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131067981","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：