![International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
@ IJTSRD | Unique Paper ID – IJTSRD30589 | Volume – 4 | Issue – 3 | March-April 2020 Page 622
grinder can used for almost 2 hrs, whereas mixer grinder
cannot be used for a long time it can be maximumtoa halfan
hour. Consumption of appliance vary according to climatic
conditions. For instance during winter seasons air
conditioner are not used. Consumption of also depend on
time of day like for night time electric bulb will be used than
during day hours. So consumption of electricity vary
according to the time of hours of day, climatic conditionsetc.
Consumption of each appliance makes the user to analysis
the consumption of electricity in building and also reduce
the consumption.
III. RELATED WORK
Energy consumption of lighting and appliances, and build-
ing controls will leads a large impact on heating, cooling and
ventilation demand. Unaware behaviour can add increase
energy consumption on building’s. User activity and
behaviour is major reasons for the larger consumption and
used for control of various devices such as artificial light,
heating, ventilation, and air conditioning . In this survey of
international intelligent buildings research efforts with the
theme of energy saving and user activity recognition. This
will devise new metrics to compare the existing studies.
From the survey, it determine theactivityand behaviour and
their effect on energy saving potential for every three main
subsystems i.e., HVAC, light, and plug loads. The important
promising and appropriate activity technologies are
recognized and approaches are examined thus it conclude
with the principles and context for energy intelligent
buildings based on user activity [1].
Demand side management (DSM) plays an important role
within the future smart grid by controlling loads in a very
smart way. DSM performed via home energy management
systems for buildings, provide many consumers enjoy
electricity price savings and utility operates at reducedpeak
demand. Evolutionaryalgorithmsare(binaryparticleswarm
optimization, genetic algorithm, and cuckoo search) DSM
model for scheduling the appliances of residential users is
presented. The model is simulated in time use pricing
environment for 3 cases: 1) traditional homes 2) smart
homes 3) smart homes with renewable energy sources.
Duplication output show that the proposed model optimally
schedules the appliances leading to electricity bill andpeaks
reductions[3].
Building’s energy consumption prediction is a major issue
and many studies have been done to improve the energy
management of buildings. Predictionof energyconsumption
in building is important for the energy to build a energy
strategy, which could be integrated to building’s energy
management system.Energy consumptionon buildingusesa
prediction model Support Vector Machine (SVM). SVM
provides a accurate way for selection of training data. The
data selection method based on Dynamic Time Warping is
used to train SVM model. In order, to enclose thermal inertia
of building, pseudo dynamic model is used that it takes into
account of information of transition of energy consumption
effects and occupancy profile. The output shows that data
selection method based on support vector machine to
predict the energy consumption of building with a high
accuracy in compare to whole data training. Data selection
method is a easy way to perform (around 8 minute training
time)in contradict to whole data training (around 31 hour
for weekend and 116 hour for working days) and admit real
control implementation for online system as well[4].
The study presents three modeling techniques for the
prediction of electricity energy consumption. The normal
regression analysis, neural network and decision tree are
considered. Model selection is completed by the root of
average squared error. In applicationtoanelectricity energy
consumption study, the choice tree and neural network
models appear to be viable alternatives to the step wise
regression model in the understanding energy consumption
patterns and predicting energy consumption levels. The
event of the information mining access for predictivemodel,
different models is in-built a unified platform: to implement
various modeling techniques, assess the performance of
various models and choose the foremost appropriate model
for future prediction[2].
Some of the challenges to predict energy utilization has
gained recognition within the residential sector because of
the significant energy consumption in recent decades. The
modeling of the residential building energy consumption is
an underdeveloped for the reduced land robust solutions
still now while this research are a has become of greater
relevance with significant advances in computation and
simulation. Such advances include the appearance of
artificial intelligence re-search in statistical model
development. The Artificial neural network has been
emerged as a basic method to handle the difficulty of non
linearity of building energy datas and also the robust
calculation of massive and dynamic data. The progress or
development, validation of such models on one amongst the
TxAIRE Research houses has been established. The TxAIRE
houses are designed to function realistic test facilities for
demonstrating new technologies. The input variables
collected from the house data are as followings the quantity
of days, outdoor temperature and radiation and also the
output variables are house and warmth pump energy
consumption. The design which relies on Leven berg-
MarquardtandOWO-Newtonalgorithmshad promising
results of coefficients of de-termination within 0.87–0.91,
which is love prior literature. Further work are explored to
develop a sturdy model for residential building
application[6]
Energy is that the lifeblood of recent societies. In past
decades, the world’s energy consumption and related to the
CO2 emissions increased rapidly because of the increases in
population and comforts of the people. Building energy
consumption prediction is crucial for the energy planning,
management, and conservation of resources. Data driven
mod-els will provide an honest approach to the energy
consumption prediction. The studies reviewed the
developed data-driven building energy consumption
prediction models, with a specific concentrate on reviewing
the scopes of prediction, the info properties and therefore
the data pre processing methods used, the machine learning
algorithms are utilized for the prediction, and performance
measures used for evaluation. supported reviews, the
present research gaps are identified and future research
directions within the area of data-driven building energy
consumption prediction are highlighted[5].
IV. PROPOSED SYSTEM
In this project, we proposea mobileapplicationthatin which
user can predict the values of electric consumption of each
appliances in a residential building. User canalsousemobile
application in order to give reviews and feedback. This uses](https://image.slidesharecdn.com/127predictionofpowerconsumptionandleakagedetection-200601114801/85/Prediction-of-Power-Consumption-and-Leakage-Detection-2-320.jpg)
![International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
@ IJTSRD | Unique Paper ID – IJTSRD30589 | Volume – 4 | Issue – 3 | March-April 2020 Page 623
previous data set of consumption and comparing it with
current data set consumption and predicts the new
consumption. The prediction model will be derived from
machine learning techniques like k means algorithm, SVM
the system provide current usage electric consumption of
each appliance can be and also predicts the consumption it
can in future.
Fig.1. A block visualization for the proposed system
CONCLUSION AND FUTURE WORK
Implemented methodology has provided very interesting
and valuable results. The load estimation process provided
good results in the analysis of the sensitivities has revealed
itself effective for finding the most relevant variables. The
overall quality of the results support the adequacy of the
adopted approach. The implemented tool both can be used
for the energy efficiency classification of buildings. The
choice of prediction model becomes the importantinfluence
factors how to improve price forecasting accuracy. Level
consumption of appliances are directly measuredsothatthe
lifestyle of the household is easier to identify, which further
assist in the forecasting. Based on this realization, rather
than serving only aggregated energy data a way is
implemented propose to in- putall availablemajorappliance
energy sequences to train the predictor. In real-world
applications, the energy consumption load in buildings
features a relationship with several underlying factors, like
temperature, humidity, work time, holidays, occupants can
provide more information about the energy consumption
variability and uncertainty. Thus, the proposed approach is
modeled to handle multiple input parameters andlargedata
non-linear prediction. If these factors considered, the
proposed model may result to better prediction accuracy.
Data acquired is processed using k means clustering using
partitioning method. Feature extracted wasmoreinorder to
predict in an accurate way. The customer could check the
predicted bill according to their usage in the previous
months. The user can check their usage in appliances and
accordingly reduce the consumption. Theproject will helpin
developing an awareness among the people towards the
usage of electricity. The user can track their usage through
smart phones. The scheme is based on machinelearningand
processing of data. All the output testing is done many times
and errors are corrected. We believe that the project will
bring a helpful hand for society and check their usage of
electricity consumption and reduce if necessary.
REFERENCE
[1] T. A. Nguyen and M. Aiello,, ““Energy intelligent
buildings based on user activity: A survey””,vol.56,pp.
244–257, Jan. 2013
[2] N. Javaid et al, (2003,““An intelligent loadmanagement
system with renewable energy integration for smart
homes””, IEEE Access, vol. 5, pp. 13587–13600, Jun.
2017.
[3] S. Paudel, P. H. Nguyen, W. L. Kling, M. Elmitri, B.
Lacarriere, and O. ‘ Le Corre,“e “Support vector
machine in prediction of buildingenergydemandusing
pseudo dynamic approach””, in Proc. 28th Int. Conf.
Efciency, Cost, Optim., Simulation Environ. Impact
Energy Syst., Pau, France, Jun. 2015.
[4] G. K. F. Tso and K. K. W. Yau,““Predicting electricity
energy consumption: A comparison of regression
analysis, decision tree and neural networks””,
Energy,vol.32,no.9,pp. 1761–1768,Sep.2007.
[5] M. A. R. Biswas, M. D. Robinson, and N.
Fumo,““Prediction of residential building energy
consumption: A neural network approach””, Energy,
vol. 117, pp. 84–92, Dec. 2016.
[6] R. K. Jain, K. M. Smith, P. J. Culligan, and J. E. Taylor,
““Forecasting energy consumption of multi-family
residential buildings using support vector regression
Investigating the impact of temporal and spatial
monitoring granularity on performance accuracy””,
Appl. Energy, vol. 123, pp. 168–178, Jun. 2014](https://image.slidesharecdn.com/127predictionofpowerconsumptionandleakagedetection-200601114801/85/Prediction-of-Power-Consumption-and-Leakage-Detection-3-320.jpg)
Prediction of Power Consumption and Leakage Detection
- 1. International Journal of Trend in Scientific Research and Development (IJTSRD) Volume 4 Issue 3, April 2020 Available Online: www.ijtsrd.com e-ISSN: 2456 – 6470 @ IJTSRD | Unique Paper ID – IJTSRD30589 | Volume – 4 | Issue – 3 | March-April 2020 Page 621 Prediction of Power Consumption and Leakage Detection Vyshnavi B1, Selma P. T1, Renin Joy1, Sheethal M. S2 1Student, 2Assistant Professor, 1,2Department of Computer Science, Sahrdaya College of Engineering and Technology, Thrissur, Kerala, India ABSTRACT Energy consumption and predictions for residential buildings play an important role within the energy management and system, as the availability and demand of energy resources is the dynamic andseasonal changes.Human beings are unaware of the cost of energy consumed by various appliancesand the energy resources available for the next generation. Each appliances in homes will consume different power consumption in different seasons. Accordingly the bill rate changes. KEYWORDS: Phone, Electric Meter, Detector How to cite this paper: Vyshnavi B | Selma P. T | Renin Joy | Sheethal M. S "Prediction of Power Consumption and Leakage Detection" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456- 6470, Volume-4 | Issue-3, April 2020, pp.621-623, URL: www.ijtsrd.com/papers/ijtsrd30589.pdf Copyright © 2020 by author(s) and International Journal ofTrendinScientific Research and Development Journal. This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (CC BY 4.0) (http://creativecommons.org/licenses/by /4.0) I. INTRODUCTION The energy resources today are closer to consumers due to sustainable energy resources and advanced technology.The flourishing concern is about the reduction of energy consumption in buildings makes it decisive to predict future energy consumption precisely using an advance predictive model. Micro grid is a power scheme that proposes closer power generation to consumers using renewable resources in rooftop panels at a local energy storages and buildings.By the utilization of renewable energy at the consumer level such as buildings, the consumption will be to get cheaper and cleaner energy resources. However, there will be some energy consumed in buildings from the local grid which has to be adjusted and predicted efficiently to reduce the consumption cost and environmental impacts. Energy consumption in build-ings accounts for a largeproportion of the primary energy worldwide and plays a vital role in the carbon emission.Hence, predictionof energyconsumptionat a building level has become a vital topicanditis necessary to develop a reliable optimization predictive model, to lower the energy costs and improve environmental buildings. Electric consumption also depend uponweatherconditions, during summer seasons cooling appliances like fans, air conditioner are used frequently, whereas during winter season it is much low. Some appliances like mixer grinder, iron box cannot be used continuously for a long time whereas appliances like refrigerator are used continuously for long time. User finds it difficult get detailed information about the consumption of electricity of each appliance. Electricity bill will give overall consumption of electricity in one or two month. In this bill consumption of different appliances is not given, it is difficult for the user to find which appliance consumed more electric power. This helps the users to create a mobile application for electric consumption. the normal user can view electriccon- sumption of each appliances. User can view electric prediction from previous data within a mobile application. User can also have ability to send complaints and its feedback. To implement a effective machine learning strategy using k means clustering and support vector machine to predict energy consumption. Differentappliance will consume different electricpower.Thesystem will power consumption of each appliance separately. The scopeforthe project is only going to get better over time and work effectively. II. PROBLEM DEFINITION In residential electricityconsumptionwill accountforlargest proportion in energy consumption. Consumption of energy used in different appliances are different. Many ap-pliances are used in residential building. Some appliances are used regularly, some are not, some are rarely used. User finds it difficult to analysis consumption of different appliances separately. Electricity bill will only giveoverall consumption of power. User is unaware of which appliance consumed more power , so user find it difficult to reduce the power consumption of appliance in residential building. In residential building appliances are used as per demand. Some appliance can be used for a long time. For instance IJTSRD30589
- 2. International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470 @ IJTSRD | Unique Paper ID – IJTSRD30589 | Volume – 4 | Issue – 3 | March-April 2020 Page 622 grinder can used for almost 2 hrs, whereas mixer grinder cannot be used for a long time it can be maximumtoa halfan hour. Consumption of appliance vary according to climatic conditions. For instance during winter seasons air conditioner are not used. Consumption of also depend on time of day like for night time electric bulb will be used than during day hours. So consumption of electricity vary according to the time of hours of day, climatic conditionsetc. Consumption of each appliance makes the user to analysis the consumption of electricity in building and also reduce the consumption. III. RELATED WORK Energy consumption of lighting and appliances, and build- ing controls will leads a large impact on heating, cooling and ventilation demand. Unaware behaviour can add increase energy consumption on building’s. User activity and behaviour is major reasons for the larger consumption and used for control of various devices such as artificial light, heating, ventilation, and air conditioning . In this survey of international intelligent buildings research efforts with the theme of energy saving and user activity recognition. This will devise new metrics to compare the existing studies. From the survey, it determine theactivityand behaviour and their effect on energy saving potential for every three main subsystems i.e., HVAC, light, and plug loads. The important promising and appropriate activity technologies are recognized and approaches are examined thus it conclude with the principles and context for energy intelligent buildings based on user activity [1]. Demand side management (DSM) plays an important role within the future smart grid by controlling loads in a very smart way. DSM performed via home energy management systems for buildings, provide many consumers enjoy electricity price savings and utility operates at reducedpeak demand. Evolutionaryalgorithmsare(binaryparticleswarm optimization, genetic algorithm, and cuckoo search) DSM model for scheduling the appliances of residential users is presented. The model is simulated in time use pricing environment for 3 cases: 1) traditional homes 2) smart homes 3) smart homes with renewable energy sources. Duplication output show that the proposed model optimally schedules the appliances leading to electricity bill andpeaks reductions[3]. Building’s energy consumption prediction is a major issue and many studies have been done to improve the energy management of buildings. Predictionof energyconsumption in building is important for the energy to build a energy strategy, which could be integrated to building’s energy management system.Energy consumptionon buildingusesa prediction model Support Vector Machine (SVM). SVM provides a accurate way for selection of training data. The data selection method based on Dynamic Time Warping is used to train SVM model. In order, to enclose thermal inertia of building, pseudo dynamic model is used that it takes into account of information of transition of energy consumption effects and occupancy profile. The output shows that data selection method based on support vector machine to predict the energy consumption of building with a high accuracy in compare to whole data training. Data selection method is a easy way to perform (around 8 minute training time)in contradict to whole data training (around 31 hour for weekend and 116 hour for working days) and admit real control implementation for online system as well[4]. The study presents three modeling techniques for the prediction of electricity energy consumption. The normal regression analysis, neural network and decision tree are considered. Model selection is completed by the root of average squared error. In applicationtoanelectricity energy consumption study, the choice tree and neural network models appear to be viable alternatives to the step wise regression model in the understanding energy consumption patterns and predicting energy consumption levels. The event of the information mining access for predictivemodel, different models is in-built a unified platform: to implement various modeling techniques, assess the performance of various models and choose the foremost appropriate model for future prediction[2]. Some of the challenges to predict energy utilization has gained recognition within the residential sector because of the significant energy consumption in recent decades. The modeling of the residential building energy consumption is an underdeveloped for the reduced land robust solutions still now while this research are a has become of greater relevance with significant advances in computation and simulation. Such advances include the appearance of artificial intelligence re-search in statistical model development. The Artificial neural network has been emerged as a basic method to handle the difficulty of non linearity of building energy datas and also the robust calculation of massive and dynamic data. The progress or development, validation of such models on one amongst the TxAIRE Research houses has been established. The TxAIRE houses are designed to function realistic test facilities for demonstrating new technologies. The input variables collected from the house data are as followings the quantity of days, outdoor temperature and radiation and also the output variables are house and warmth pump energy consumption. The design which relies on Leven berg- MarquardtandOWO-Newtonalgorithmshad promising results of coefficients of de-termination within 0.87–0.91, which is love prior literature. Further work are explored to develop a sturdy model for residential building application[6] Energy is that the lifeblood of recent societies. In past decades, the world’s energy consumption and related to the CO2 emissions increased rapidly because of the increases in population and comforts of the people. Building energy consumption prediction is crucial for the energy planning, management, and conservation of resources. Data driven mod-els will provide an honest approach to the energy consumption prediction. The studies reviewed the developed data-driven building energy consumption prediction models, with a specific concentrate on reviewing the scopes of prediction, the info properties and therefore the data pre processing methods used, the machine learning algorithms are utilized for the prediction, and performance measures used for evaluation. supported reviews, the present research gaps are identified and future research directions within the area of data-driven building energy consumption prediction are highlighted[5]. IV. PROPOSED SYSTEM In this project, we proposea mobileapplicationthatin which user can predict the values of electric consumption of each appliances in a residential building. User canalsousemobile application in order to give reviews and feedback. This uses
- 3. International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470 @ IJTSRD | Unique Paper ID – IJTSRD30589 | Volume – 4 | Issue – 3 | March-April 2020 Page 623 previous data set of consumption and comparing it with current data set consumption and predicts the new consumption. The prediction model will be derived from machine learning techniques like k means algorithm, SVM the system provide current usage electric consumption of each appliance can be and also predicts the consumption it can in future. Fig.1. A block visualization for the proposed system CONCLUSION AND FUTURE WORK Implemented methodology has provided very interesting and valuable results. The load estimation process provided good results in the analysis of the sensitivities has revealed itself effective for finding the most relevant variables. The overall quality of the results support the adequacy of the adopted approach. The implemented tool both can be used for the energy efficiency classification of buildings. The choice of prediction model becomes the importantinfluence factors how to improve price forecasting accuracy. Level consumption of appliances are directly measuredsothatthe lifestyle of the household is easier to identify, which further assist in the forecasting. Based on this realization, rather than serving only aggregated energy data a way is implemented propose to in- putall availablemajorappliance energy sequences to train the predictor. In real-world applications, the energy consumption load in buildings features a relationship with several underlying factors, like temperature, humidity, work time, holidays, occupants can provide more information about the energy consumption variability and uncertainty. Thus, the proposed approach is modeled to handle multiple input parameters andlargedata non-linear prediction. If these factors considered, the proposed model may result to better prediction accuracy. Data acquired is processed using k means clustering using partitioning method. Feature extracted wasmoreinorder to predict in an accurate way. The customer could check the predicted bill according to their usage in the previous months. The user can check their usage in appliances and accordingly reduce the consumption. Theproject will helpin developing an awareness among the people towards the usage of electricity. The user can track their usage through smart phones. The scheme is based on machinelearningand processing of data. All the output testing is done many times and errors are corrected. We believe that the project will bring a helpful hand for society and check their usage of electricity consumption and reduce if necessary. REFERENCE [1] T. A. Nguyen and M. Aiello,, ““Energy intelligent buildings based on user activity: A survey””,vol.56,pp. 244–257, Jan. 2013 [2] N. Javaid et al, (2003,““An intelligent loadmanagement system with renewable energy integration for smart homes””, IEEE Access, vol. 5, pp. 13587–13600, Jun. 2017. [3] S. Paudel, P. H. Nguyen, W. L. Kling, M. Elmitri, B. Lacarriere, and O. ‘ Le Corre,“e “Support vector machine in prediction of buildingenergydemandusing pseudo dynamic approach””, in Proc. 28th Int. Conf. Efciency, Cost, Optim., Simulation Environ. Impact Energy Syst., Pau, France, Jun. 2015. [4] G. K. F. Tso and K. K. W. Yau,““Predicting electricity energy consumption: A comparison of regression analysis, decision tree and neural networks””, Energy,vol.32,no.9,pp. 1761–1768,Sep.2007. [5] M. A. R. Biswas, M. D. Robinson, and N. Fumo,““Prediction of residential building energy consumption: A neural network approach””, Energy, vol. 117, pp. 84–92, Dec. 2016. [6] R. K. Jain, K. M. Smith, P. J. Culligan, and J. E. Taylor, ““Forecasting energy consumption of multi-family residential buildings using support vector regression Investigating the impact of temporal and spatial monitoring granularity on performance accuracy””, Appl. Energy, vol. 123, pp. 168–178, Jun. 2014