The present study considers decreasing prediction error for the types of time series and the unce... more The present study considers decreasing prediction error for the types of time series and the uncertainty in estimation parameters, improving the structure of the deep neural network and increasing response speed in the proposed neural network method; besides, the competitive performance and the collaboration among the neurons of deep neural network are also increased. Selected data is related to Qeshm weather (suitable weather conditions to study our purpose) prediction during 2016 onwards. In this study, for the purpose of analyzing the prediction issue of power consumption of domestic expenses in the indefinite and severe fluctuation mode, we decided to combine two methods of Long Short-Term Memory and Convolutional Neural Network. For the training of the deep network, the BP algorithm is used. The results indicate that Gated Recurrent Unit networks compared other models (MLP, CNN and DNN) produce more realistic results, and also two-way networks obtained better results on test data compared Long Short-Term Memory networks. RMSE prediction are more realistic than the LSTM model on test and training data against the significant data. A GRU network has two gates of rt readjust and Zt forgetting, which helps to assure that long term dependencies of gradient fading will not occur.
Wind speed is one of the most vital, imperative meteorological parameters, thus the prediction of... more Wind speed is one of the most vital, imperative meteorological parameters, thus the prediction of which is of fundamental importance in the studies related to energy management, building construction, damages caused by strong winds, aquatic needs of power plants, the prevalence and spread of diseases, snowmelt, and air pollution. Due to the discrete and nonlinear structure of wind speed, wind speed forecasting at regular intervals is a crucial problem. In this regard, a wide variety of prediction methods have been applied. So far, many activities have been done in order to make optimal use of renewable energy sources such as wind, which have led to the present diverse types of wind speed and strength measuring methods in the various geographical locations. In this paper, a novel forecasting model based on hybrid neural networks (HNNs) and wavelet packet decomposition (WPD) processor has been proposed to predict wind speed. Considering this scenario, the accuracy of the proposed method is compared with other wind speed prediction methods to ensure performance improvement.
This work was conducted to design a combined cooling, heating, and power (CCHP) system with photo... more This work was conducted to design a combined cooling, heating, and power (CCHP) system with photovoltaic energy which provides simultaneous generation of electricity, heat, and cold for a high-rise office building (23 floors) in the city of Mashhad in Iran. Our strategy was to supply load electric, thermal, and refrigeration with the help of solar energy. In addition, its superiority over other systems was evaluated. Analysis and study of solar radiation and the maximum level of solar panels use, according to the architectural plan, were carried out at the project site. The analysis of shadow points, the use of inverters and electrical detectors to increase the maximum solar power, and its cost-effectiveness were carefully studied via PVSOL software. Additionally, the amount of heat, cold, and electricity consumption was accurately calculated according to international standards and utilizing HAP software. The criteria for saving on the initial cost reduction, carbon dioxide emission reduction, operating cost reduction, payback period, revenue, and the minimum life expectancy of the equipment compared to those in other methods were also evaluated. The results obtained from the designed system of simultaneous generation of electricity, heat, and refrigeration, which combines gas microturbines as the primary stimulus, a combination of absorption and compression chiller to provide refrigeration load, a boiler for auxiliary heat load, and a thermal photovoltaic system to produce both electric and thermal loads, were finally revealed. This is believed to be a cost-effective strategy for high-rise residential or commercial buildings with a geographical location like that of Mashhad. Based on the electricity sales to the grid, with the rate of increase in inflation in electricity tariffs, this design in the Mashhad project was estimated to have an annual income of 166.676 thousand
In recent years due to increased competition between companies in the services sector, predict ch... more In recent years due to increased competition between companies in the services sector, predict churn customer in order to retain customers is so important. The impact of brand loyalty and customer churn in an organization as well as the difficulty of attracting a new customer per lost customer is very painful for organizations. Obtaining a predictive model customer behaviour to plan for and deal with such cases, can be very helpful. Employee churn or loss of staff will be close to the customer churn, but the impact of losing a major customer for organization certainly will be more painful (because organization do not have physical sense to losing their employees) while the consequences of finding well employees instead of missed employees, As well as the cost of in-service training that should be given to new employees could be one of the issues that each organization would be sensitive to losing its human resources.
A Bayesian classifier is one of the most widely used classifiers which possess several properties... more A Bayesian classifier is one of the most widely used classifiers which possess several properties that make it surprisingly useful and accurate. It is illustrated that performance of Bayesian learning in some cases is comparable with neural networks and decision trees. Bayesian theorem suggests a straight forward process which is not based on search methods. This is the major point which satisfies the marvelous time complexity of Bayesian classifier. At the other hand, constructing phase of fuzzy intrusion detection systems suffer from time consuming processes which are based on search methods. In this paper we propose a novel method to accelerate such processes using Bayesian inference. Experimental results show meaningful time reduction.
Artificial neural networks are mathematical models that have been inspired by the human nervous s... more Artificial neural networks are mathematical models that have been inspired by the human nervous system and brain. In this study, the purpose is establishing a method to predict the cost of tractor repair and maintenance more accurately. In this paper, the multi-layer neural network with Feed Forward Backpropagation training algorithm (FFBP) has been used to predict repair and maintenance costs of tractor. In addition, 60 real data from two-wheel drive tractors existing in Razavi agro-industry in Iran have been used to train and test the network. The appropriate parameters for network training have been selected through error test on data. In this study, the performance of Backpropagate Declining Learning Rate Factor algorithm (BDLRF) has been compared with Feed-Forward Backpropagation algorithm (FFBP), with error criteria (MAPE, MSE, and RMSE) and the result shows that training Feed Forward Backpropagation algorithm (FFBP) surpasses the (BDLRF) algorithm in predicting tractor repair and maintenance costs by using separate networks rather than a single network.
In a competitive electricity market, price forecasts are important for market participants. Howev... more In a competitive electricity market, price forecasts are important for market participants. However, electricity price is a complex signal due to its nonlinearity, nonstationarity, and time variant behavior. In spite of much research in this area, more accurate and robust price forecast methods are still required. In this paper, a combination of a feature selection technique and cascaded neuro-evolutionary algorithm (CNEA) is proposed for this purpose. The feature selection method is an improved version of the mutual information (MI) technique. The CNEA is composed of cascaded forecasters where each forecaster consists of a neural network (NN) and an evolutionary algorithm (EA). An iterative search procedure is also incorporated in our solution strategy to fine-tune the adjustable parameters of both the MI technique and CNEA. The price forecast accuracy of the proposed method is evaluated by means of real data from the Pennsylvania-New Jersey-Maryland (PJM) and Spanish electricity markets. The method is also compared with some of the most recent price forecast techniques. Index Terms-Cascaded neuro-evolutionary algorithm (CNEA), iterative search procedure, mutual information (MI), price forecast.
Short-term electricity load and price forecasting is an important issue in competitive electricit... more Short-term electricity load and price forecasting is an important issue in competitive electricity markets. In this paper, we propose a new direct time series forecasting method based on clustering and next symbol prediction. First, the cluster label sequence is obtained from time series clustering. Then a lossless compression algorithm of prediction by partial match version C coder (PPMC) is applied on this obtained discrete cluster label sequence to predict the next cluster label. Finally, the whole time series values of one-step-ahead can be directly forecast from the predicted cluster label. The proposed method is evaluated on electricity time series datasets, and the numerical experiments show that the proposed method can achieve promising results in day-ahead electricity load and price forecasting.
Social networks are streaming, and diverse including a wide range of edges so that they are conti... more Social networks are streaming, and diverse including a wide range of edges so that they are continuously evolved over time and are formed by the activities such as tweets and emails among the users; each activity, adds an edge to the network graph. Despite their popularity, the dynamically and large size of most social networks make it difficult or impossible to study the entire networks. This paper proposes a sampling algorithm that is equipped with an evaluator unit for analyzing the edges and a set of simple fixed structure learning automata. The evaluator unit evaluates each edge and then decides whether the edge and the corresponding node should be added to the sample set. In the proposed algorithm, each main activity graph node is equipped with a simple learning automaton. The developed algorithm is compared with the best current sampling algorithm reported in the Kolmogorov-Smirnov test, and the normalized L1 and L2 distances in real networks and synthetic networks are presented as a sequence of edges. The experimental results obtained show the superiority of the proposed algorithm.
Energy is one of the most important topics in the world today and is considered as one of the mos... more Energy is one of the most important topics in the world today and is considered as one of the most effective factors for the development of countries. Due to the limitation of non-renewable energy sources and undesirable effects of consuming these resources on the environment, the strategy of countries has changed towards the use of renewable energy. Renewable energy sources do not decrease over time and operate independently of price fluctuations and are more available, thus being able to play a greater role in modern power systems. Therefore, the optimal location and use of these resources will have an impact on modifying the parameters of the power grid. In this paper an analytical approach for optimal placement and sizing of distributed generation (DG) in power distribution networks to minimize the power loss, bus voltage limits, DG capacity limits, current limits, and DG penetration limit. In the first step, determines the DG capacity causing maximum benefit at different buses,...
International Journal of Electrical Power & Energy Systems, Nov 1, 2008
In a competitive electricity market, forecast of energy prices is a key information for the marke... more In a competitive electricity market, forecast of energy prices is a key information for the market participants. However, price signal usually has a complex behavior due to its nonlinearity, nonstationarity, and time variancy. In spite of all performed researches on this area in the recent years, there is still an essential need for more accurate and robust price forecast methods. In this paper, a combination of wavelet transform (WT) and a hybrid forecast method is proposed for this purpose. The hybrid method is composed of cascaded forecasters where each forecaster consists of a neural network (NN) and an evolutionary algorithms (EA). Both time domain and wavelet domain features are considered in a mixed data model for price forecast, in which the candidate input variables are refined by a feature selection technique. The adjustable parameters of the whole method are fine-tuned by a cross-validation technique. The proposed method is examined on PJM electricity market and compared with some of the most recent price forecast methods.
Reliability and accessibility of distribution systems are important goals that have significant i... more Reliability and accessibility of distribution systems are important goals that have significant impacts on the costs. The proper strategy of maintenance based on components arrangement and assets is the best way to reach these goals. This strategy is a kind of uses reliability-centered maintenance (RCM). Due to the limited maintenance budget, performing maintenance activities for all components of the system is neither possible nor logical. So most of the resources should be allocated to the most critical and important components. This paper presents a novel analytical method of prioritization of distribution systems’ components by introducing a new weighted cumulative Reliability-based diagnostic importance factor. This new factor includes different reliability indexes in form of diagnostic factors and will show that the order of components obtained by this method is better than another method in saving the budget and providing reliability of the system. The process of decision-mak...
The present study considers decreasing prediction error for the types of time series and the unce... more The present study considers decreasing prediction error for the types of time series and the uncertainty in estimation parameters, improving the structure of the deep neural network and increasing response speed in the proposed neural network method; besides, the competitive performance and the collaboration among the neurons of deep neural network are also increased. Selected data is related to Qeshm weather (suitable weather conditions to study our purpose) prediction during 2016 onwards. In this study, for the purpose of analyzing the prediction issue of power consumption of domestic expenses in the indefinite and severe fluctuation mode, we decided to combine two methods of Long Short-Term Memory and Convolutional Neural Network. For the training of the deep network, the BP algorithm is used. The results indicate that Gated Recurrent Unit networks compared other models (MLP, CNN and DNN) produce more realistic results, and also two-way networks obtained better results on test data compared Long Short-Term Memory networks. RMSE prediction are more realistic than the LSTM model on test and training data against the significant data. A GRU network has two gates of rt readjust and Zt forgetting, which helps to assure that long term dependencies of gradient fading will not occur.
Wind speed is one of the most vital, imperative meteorological parameters, thus the prediction of... more Wind speed is one of the most vital, imperative meteorological parameters, thus the prediction of which is of fundamental importance in the studies related to energy management, building construction, damages caused by strong winds, aquatic needs of power plants, the prevalence and spread of diseases, snowmelt, and air pollution. Due to the discrete and nonlinear structure of wind speed, wind speed forecasting at regular intervals is a crucial problem. In this regard, a wide variety of prediction methods have been applied. So far, many activities have been done in order to make optimal use of renewable energy sources such as wind, which have led to the present diverse types of wind speed and strength measuring methods in the various geographical locations. In this paper, a novel forecasting model based on hybrid neural networks (HNNs) and wavelet packet decomposition (WPD) processor has been proposed to predict wind speed. Considering this scenario, the accuracy of the proposed method is compared with other wind speed prediction methods to ensure performance improvement.
This work was conducted to design a combined cooling, heating, and power (CCHP) system with photo... more This work was conducted to design a combined cooling, heating, and power (CCHP) system with photovoltaic energy which provides simultaneous generation of electricity, heat, and cold for a high-rise office building (23 floors) in the city of Mashhad in Iran. Our strategy was to supply load electric, thermal, and refrigeration with the help of solar energy. In addition, its superiority over other systems was evaluated. Analysis and study of solar radiation and the maximum level of solar panels use, according to the architectural plan, were carried out at the project site. The analysis of shadow points, the use of inverters and electrical detectors to increase the maximum solar power, and its cost-effectiveness were carefully studied via PVSOL software. Additionally, the amount of heat, cold, and electricity consumption was accurately calculated according to international standards and utilizing HAP software. The criteria for saving on the initial cost reduction, carbon dioxide emission reduction, operating cost reduction, payback period, revenue, and the minimum life expectancy of the equipment compared to those in other methods were also evaluated. The results obtained from the designed system of simultaneous generation of electricity, heat, and refrigeration, which combines gas microturbines as the primary stimulus, a combination of absorption and compression chiller to provide refrigeration load, a boiler for auxiliary heat load, and a thermal photovoltaic system to produce both electric and thermal loads, were finally revealed. This is believed to be a cost-effective strategy for high-rise residential or commercial buildings with a geographical location like that of Mashhad. Based on the electricity sales to the grid, with the rate of increase in inflation in electricity tariffs, this design in the Mashhad project was estimated to have an annual income of 166.676 thousand
In recent years due to increased competition between companies in the services sector, predict ch... more In recent years due to increased competition between companies in the services sector, predict churn customer in order to retain customers is so important. The impact of brand loyalty and customer churn in an organization as well as the difficulty of attracting a new customer per lost customer is very painful for organizations. Obtaining a predictive model customer behaviour to plan for and deal with such cases, can be very helpful. Employee churn or loss of staff will be close to the customer churn, but the impact of losing a major customer for organization certainly will be more painful (because organization do not have physical sense to losing their employees) while the consequences of finding well employees instead of missed employees, As well as the cost of in-service training that should be given to new employees could be one of the issues that each organization would be sensitive to losing its human resources.
A Bayesian classifier is one of the most widely used classifiers which possess several properties... more A Bayesian classifier is one of the most widely used classifiers which possess several properties that make it surprisingly useful and accurate. It is illustrated that performance of Bayesian learning in some cases is comparable with neural networks and decision trees. Bayesian theorem suggests a straight forward process which is not based on search methods. This is the major point which satisfies the marvelous time complexity of Bayesian classifier. At the other hand, constructing phase of fuzzy intrusion detection systems suffer from time consuming processes which are based on search methods. In this paper we propose a novel method to accelerate such processes using Bayesian inference. Experimental results show meaningful time reduction.
Artificial neural networks are mathematical models that have been inspired by the human nervous s... more Artificial neural networks are mathematical models that have been inspired by the human nervous system and brain. In this study, the purpose is establishing a method to predict the cost of tractor repair and maintenance more accurately. In this paper, the multi-layer neural network with Feed Forward Backpropagation training algorithm (FFBP) has been used to predict repair and maintenance costs of tractor. In addition, 60 real data from two-wheel drive tractors existing in Razavi agro-industry in Iran have been used to train and test the network. The appropriate parameters for network training have been selected through error test on data. In this study, the performance of Backpropagate Declining Learning Rate Factor algorithm (BDLRF) has been compared with Feed-Forward Backpropagation algorithm (FFBP), with error criteria (MAPE, MSE, and RMSE) and the result shows that training Feed Forward Backpropagation algorithm (FFBP) surpasses the (BDLRF) algorithm in predicting tractor repair and maintenance costs by using separate networks rather than a single network.
In a competitive electricity market, price forecasts are important for market participants. Howev... more In a competitive electricity market, price forecasts are important for market participants. However, electricity price is a complex signal due to its nonlinearity, nonstationarity, and time variant behavior. In spite of much research in this area, more accurate and robust price forecast methods are still required. In this paper, a combination of a feature selection technique and cascaded neuro-evolutionary algorithm (CNEA) is proposed for this purpose. The feature selection method is an improved version of the mutual information (MI) technique. The CNEA is composed of cascaded forecasters where each forecaster consists of a neural network (NN) and an evolutionary algorithm (EA). An iterative search procedure is also incorporated in our solution strategy to fine-tune the adjustable parameters of both the MI technique and CNEA. The price forecast accuracy of the proposed method is evaluated by means of real data from the Pennsylvania-New Jersey-Maryland (PJM) and Spanish electricity markets. The method is also compared with some of the most recent price forecast techniques. Index Terms-Cascaded neuro-evolutionary algorithm (CNEA), iterative search procedure, mutual information (MI), price forecast.
Short-term electricity load and price forecasting is an important issue in competitive electricit... more Short-term electricity load and price forecasting is an important issue in competitive electricity markets. In this paper, we propose a new direct time series forecasting method based on clustering and next symbol prediction. First, the cluster label sequence is obtained from time series clustering. Then a lossless compression algorithm of prediction by partial match version C coder (PPMC) is applied on this obtained discrete cluster label sequence to predict the next cluster label. Finally, the whole time series values of one-step-ahead can be directly forecast from the predicted cluster label. The proposed method is evaluated on electricity time series datasets, and the numerical experiments show that the proposed method can achieve promising results in day-ahead electricity load and price forecasting.
Social networks are streaming, and diverse including a wide range of edges so that they are conti... more Social networks are streaming, and diverse including a wide range of edges so that they are continuously evolved over time and are formed by the activities such as tweets and emails among the users; each activity, adds an edge to the network graph. Despite their popularity, the dynamically and large size of most social networks make it difficult or impossible to study the entire networks. This paper proposes a sampling algorithm that is equipped with an evaluator unit for analyzing the edges and a set of simple fixed structure learning automata. The evaluator unit evaluates each edge and then decides whether the edge and the corresponding node should be added to the sample set. In the proposed algorithm, each main activity graph node is equipped with a simple learning automaton. The developed algorithm is compared with the best current sampling algorithm reported in the Kolmogorov-Smirnov test, and the normalized L1 and L2 distances in real networks and synthetic networks are presented as a sequence of edges. The experimental results obtained show the superiority of the proposed algorithm.
Energy is one of the most important topics in the world today and is considered as one of the mos... more Energy is one of the most important topics in the world today and is considered as one of the most effective factors for the development of countries. Due to the limitation of non-renewable energy sources and undesirable effects of consuming these resources on the environment, the strategy of countries has changed towards the use of renewable energy. Renewable energy sources do not decrease over time and operate independently of price fluctuations and are more available, thus being able to play a greater role in modern power systems. Therefore, the optimal location and use of these resources will have an impact on modifying the parameters of the power grid. In this paper an analytical approach for optimal placement and sizing of distributed generation (DG) in power distribution networks to minimize the power loss, bus voltage limits, DG capacity limits, current limits, and DG penetration limit. In the first step, determines the DG capacity causing maximum benefit at different buses,...
International Journal of Electrical Power & Energy Systems, Nov 1, 2008
In a competitive electricity market, forecast of energy prices is a key information for the marke... more In a competitive electricity market, forecast of energy prices is a key information for the market participants. However, price signal usually has a complex behavior due to its nonlinearity, nonstationarity, and time variancy. In spite of all performed researches on this area in the recent years, there is still an essential need for more accurate and robust price forecast methods. In this paper, a combination of wavelet transform (WT) and a hybrid forecast method is proposed for this purpose. The hybrid method is composed of cascaded forecasters where each forecaster consists of a neural network (NN) and an evolutionary algorithms (EA). Both time domain and wavelet domain features are considered in a mixed data model for price forecast, in which the candidate input variables are refined by a feature selection technique. The adjustable parameters of the whole method are fine-tuned by a cross-validation technique. The proposed method is examined on PJM electricity market and compared with some of the most recent price forecast methods.
Reliability and accessibility of distribution systems are important goals that have significant i... more Reliability and accessibility of distribution systems are important goals that have significant impacts on the costs. The proper strategy of maintenance based on components arrangement and assets is the best way to reach these goals. This strategy is a kind of uses reliability-centered maintenance (RCM). Due to the limited maintenance budget, performing maintenance activities for all components of the system is neither possible nor logical. So most of the resources should be allocated to the most critical and important components. This paper presents a novel analytical method of prioritization of distribution systems’ components by introducing a new weighted cumulative Reliability-based diagnostic importance factor. This new factor includes different reliability indexes in form of diagnostic factors and will show that the order of components obtained by this method is better than another method in saving the budget and providing reliability of the system. The process of decision-mak...
Uploads
Papers by Farshid Keynia