scholarly journals Deep Learning with Loss Ensembles for Solar Power Prediction in Smart Cities

Smart Cities ◽  
2020 ◽  
Vol 3 (3) ◽  
pp. 842-852
Author(s):  
Moein Hajiabadi ◽  
Mahdi Farhadi ◽  
Vahide Babaiyan ◽  
Abouzar Estebsari
Keyword(s):  
Power Generation ◽  
Deep Learning ◽  
Large Scale ◽  
Smart Cities ◽  
Energy Generation ◽  
Loss Functions ◽  
Support Vector ◽  
Pv System ◽  
Pv Systems ◽  
Pv Power Generation

The demand for renewable energy generation, especially photovoltaic (PV) power generation, has been growing over the past few years. However, the amount of generated energy by PV systems is highly dependent on weather conditions. Therefore, accurate forecasting of generated PV power is of importance for large-scale deployment of PV systems. Recently, machine learning (ML) methods have been widely used for PV power generation forecasting. A variety of these techniques, including artificial neural networks (ANNs), ridge regression, K-nearest neighbour (kNN) regression, decision trees, support vector regressions (SVRs) have been applied for this purpose and achieved good performance. In this paper, we briefly review the most recent ML techniques for PV energy generation forecasting and propose a new regression technique to automatically predict a PV system’s output based on historical input parameters. More specifically, the proposed loss function is a combination of three well-known loss functions: Correntropy, Absolute and Square Loss which encourages robustness and generalization jointly. We then integrate the proposed objective function into a Deep Learning model to predict a PV system’s output. By doing so, both the coefficients of loss functions and weight parameters of the ANN are learned jointly via back propagation. We investigate the effectiveness of the proposed method through comprehensive experiments on real data recorded by a real PV system. The experimental results confirm that our method outperforms the state-of-the-art ML methods for PV energy generation forecasting.

scholarly journals Influence of Parasitic Parameters on DC–DC Converters and Their Method of Suppression in High Frequency Link 35 kV PV Systems

Energies ◽  
2019 ◽  
Vol 12 (19) ◽  
pp. 3743
Author(s):  
Rui Li ◽  
Fangyuan Shi ◽  
Xu Cai ◽  
Haibo Xu
Keyword(s):  
High Frequency ◽  
Large Scale ◽  
Pv System ◽  
Zero Voltage Switching ◽  
Pv Systems ◽  
Magnetic Inductance ◽  
Parasitic Parameters ◽  
Saturable Inductor ◽  
Secondary Side ◽  
Pv Power Generation

Photovoltaic (PV) power generation has shown a trend towards large-scale medium- or high-voltage integration in recent years. The development of high-frequency link PV systems is necessary for the further improvement of system efficiency and the reduction of system cost. In the system, high-frequency high-step-up ratio LLC converters are one of the most important parts. However, the parasitic parameters of devices lead to a loss of zero-voltage switching (ZVS) in the LLC converter, greatly reducing the efficiency of the system, especially in such a high-frequency application. In this paper, a high-frequency link 35 kV PV system is presented. To suppress the influences of parasitic parameters in the LLC converter in the 35 kV PV system, the influence of parasitic parameters on ZVS is analyzed and expounded. Then, a suppression method is proposed to promote the realization of ZVS. This method adds a saturable inductor on the secondary side to achieve ZVS. The saturable inductor can effectively prevent the parasitic elements of the secondary side from participating in the resonance of the primary side. The experimental results show that this method achieves a higher efficiency than the traditional method by reducing the magnetic inductance.

Crowdsourcing Incident Information for Emergency Response using Open Data Sources in Smart Cities

2018 ◽  
Vol 2672 (1) ◽  
pp. 198-208 ◽  
Author(s):  
Fan Zuo ◽  
Abdullah Kurkcu ◽  
Kaan Ozbay ◽  
Jingqin Gao
Keyword(s):  
Social Media ◽  
Prior Knowledge ◽  
Emergency Response ◽  
Large Scale ◽  
Latent Dirichlet Allocation ◽  
Smart Cities ◽  
Hurricane Sandy ◽  
Support Vector ◽  
Twitter Data ◽  
Emergency Events

Emergency events affect human security and safety as well as the integrity of the local infrastructure. Emergency response officials are required to make decisions using limited information and time. During emergency events, people post updates to social media networks, such as tweets, containing information about their status, help requests, incident reports, and other useful information. In this research project, the Latent Dirichlet Allocation (LDA) model is used to automatically classify incident-related tweets and incident types using Twitter data. Unlike the previous social media information models proposed in the related literature, the LDA is an unsupervised learning model which can be utilized directly without prior knowledge and preparation for data in order to save time during emergencies. Twitter data including messages and geolocation information during two recent events in New York City, the Chelsea explosion and Hurricane Sandy, are used as two case studies to test the accuracy of the LDA model for extracting incident-related tweets and labeling them by incident type. Results showed that the model could extract emergency events and classify them for both small and large-scale events, and the model’s hyper-parameters can be shared in a similar language environment to save model training time. Furthermore, the list of keywords generated by the model can be used as prior knowledge for emergency event classification and training of supervised classification models such as support vector machine and recurrent neural network.

scholarly journals A Clonal Selection Optimization System for Multiparty Secure Computing

Complexity ◽  
2021 ◽  
Vol 2021 ◽  
pp. 1-14
Author(s):  
Minyu Shi ◽  
Yongting Zhang ◽  
Huanhuan Wang ◽  
Junfeng Hu ◽  
Xiang Wu
Keyword(s):  
Deep Learning ◽  
Large Scale ◽  
Differential Privacy ◽  
Clonal Selection ◽  
Smart Cities ◽  
Model Performance ◽  
Processing System ◽  
Optimization System ◽  
Learning Scheme ◽  
Training Efficiency

The innovation of the deep learning modeling scheme plays an important role in promoting the research of complex problems handled with artificial intelligence in smart cities and the development of the next generation of information technology. With the widespread use of smart interactive devices and systems, the exponential growth of data volume and the complex modeling requirements increase the difficulty of deep learning modeling, and the classical centralized deep learning modeling scheme has encountered bottlenecks in the improvement of model performance and the diversification of smart application scenarios. The parallel processing system in deep learning links the virtual information space with the physical world, although the distributed deep learning research has become a crucial concern with its unique advantages in training efficiency, and improving the availability of trained models and preventing privacy disclosure are still the main challenges faced by related research. To address these above issues in distributed deep learning, this research developed a clonal selective optimization system based on the federated learning framework for the model training process involving large-scale data. This system adopts the heuristic clonal selective strategy in local model optimization and optimizes the effect of federated training. First of all, this process enhances the adaptability and robustness of the federated learning scheme and improves the modeling performance and training efficiency. Furthermore, this research attempts to improve the privacy security defense capability of the federated learning scheme for big data through differential privacy preprocessing. The simulation results show that the proposed clonal selection optimization system based on federated learning has significant optimization ability on model basic performance, stability, and privacy.

scholarly journals State of the Art Survey of Deep Learning and Machine Learning Models for Smart Cities and Urban Sustainability

2020 ◽  
Author(s):  
Saeed Nosratabadi ◽  
Amir Mosavi ◽  
Ramin Keivani ◽  
Sina Faizollahzadeh Ardabili ◽  
Farshid Aram
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
State Of The Art ◽  
Smart Cities ◽  
Model Development ◽  
Urban Sustainability ◽  
Urban Transport ◽  
Support Vector ◽  
Neuro Fuzzy ◽  
Vector Machines

Deep learning (DL) and machine learning (ML) methods have recently contributed to the advancement of models in the various aspects of prediction, planning, and uncertainty analysis of smart cities and urban development. This paper presents the state of the art of DL and ML methods used in this realm. Through a novel taxonomy, the advances in model development and new application domains in urban sustainability and smart cities are presented. Findings reveal that five DL and ML methods have been most applied to address the different aspects of smart cities. These are artificial neural networks; support vector machines; decision trees; ensembles, Bayesians, hybrids, and neuro-fuzzy; and deep learning. It is also disclosed that energy, health, and urban transport are the main domains of smart cities that DL and ML methods contributed in to address their problems.

scholarly journals Nanostructures GaAs Solar Cells

2020 ◽  
Author(s):  
Mohammad Nur E Alam ◽  
Narottam Das
Keyword(s):  
Power Generation ◽  
Solar Cells ◽  
Conversion Efficiency ◽  
Human Mobility ◽  
Incident Light ◽  
Front Surface ◽  
Pv Systems ◽  
Energy Demands ◽  
Power Generation Technology ◽  
Pv Power Generation

At present, the world is now passing a very far different time than normal situation due to the COVID-19 pandemic crisis. The global life-style and human civilization is currently progressing with down-stream that affecting almost every sectors necessary for human civilizations except the current environmental situation. To control the COVID-19 spreading, most of the countries are following lockdown process that reduces human mobility, thus reducing the CO2 emission to the environment. Though the COVID-19 pandemic is a blessing for the present environment, however, the post-COVID world will face a massive thrust of energy and only conventional energy resources may not be enough to mitigate the energy demands. Solar power generation technology mainly the photovoltaic (PV) systems and their advancement can be the leading possibilities to minimize the gap between the power demand and generation. It is now time to think how we can improve the PV power generation in future and the post-COVID world. In this encyclopaedia communication, we report on Nano-technological approach to improve the conversion efficiency of GaAs solar cells. We have designed and optimized several types of nano-structured assemblies that can be implemented to reduce the front surface incident light reflection losses thus can assist to improve the conversion efficiency of GaAs solar cells.

scholarly journals Algorithm Study on Improving Application Efficiency of Independent Solar Power Supply System

2016 ◽  
Vol 3 (1) ◽  
pp. 5
Author(s):  
Jigang Cao
Keyword(s):  
Power Generation ◽  
Maximum Power Point Tracking ◽  
Maximum Power ◽  
Maximum Power Point ◽  
Pv Systems ◽  
Point Tracking ◽  
Power Point Tracking ◽  
Important Means ◽  
Power Point ◽  
Pv Power Generation

<p class="p1"><span class="s1">With the development of photovoltaic (PV) technologies, applications of photovoltaic have grown rapidly, indicating that the photovoltaic are attractive to produce environmentally benign electricity for diversified purposes. In order to maximize the use of solar energy, this thesis focuses on the PV power generation systems, which includes modeling of PV systems, maximum power point tracking (MPPT) methods for PV arrays. </span><span class="s1">Maximum Power Point Tracking (MPPT) method is an important means to improve the system efficiency of PV power generation system. MPPT theory and various MPPT algorithms are introduced in the literature. Based on those researches, this thesis proposes a novel implementation of an adaptive duty cycle P&amp;O algorithm that can reduce the main drawbacks commonly related to the traditional P&amp;O method.</span></p>

Optimal Penetration of Photovoltaic Systems in Distribution Networks

2013 ◽  
Vol 479-480 ◽  
pp. 590-594
Author(s):  
Wei Lin Hsieh ◽  
Chia Hung Lin ◽  
Chao Shun Chen ◽  
Cheng Ting Hsu ◽  
Chin Ying Ho ◽  
...  
Keyword(s):  
Power Generation ◽  
Solar Energy ◽  
Voltage Control ◽  
Distribution Networks ◽  
Solar Irradiation ◽  
Total Power ◽  
Pv System ◽  
Pv Systems ◽  
Power Injection ◽  
Distribution Feeder

The penetration level of a PV system is often limited due to the violation of voltage variation introduced by the large intermittent power generation. This paper discusses the use of an active power curtailment strategy to reduce PV power injection during peak solar irradiation to prevent voltage violation so that the PV penetration level of a distribution feeder can be increased to fully utilize solar energy. When using the proposed voltage control scheme for limiting PV power injection into the study distribution feeder during high solar irradiation periods, the total power generation and total energy delivered by the PV system over a 1-year period are determined according to the annual duration of solar irradiation. With the proposed voltage control to perform the partial generation rejection of PV systems, the optimal installation capacity of PV systems can be determined by maximizing the net present value of the system so that better cost effectiveness of the PV project and better utilization of solar energy can be obtained.

scholarly journals Large Photovoltaic Power Plants Integration: A Review of Challenges and Solutions

Energies ◽  
2019 ◽  
Vol 12 (19) ◽  
pp. 3798 ◽  
Author(s):  
Mansouri ◽  
Lashab ◽  
Sera ◽  
Guerrero ◽  
Cherif
Keyword(s):  
System Integration ◽  
Power Plants ◽  
Large Scale ◽  
System Stability ◽  
Pv System ◽  
Pv Systems ◽  
Large Scale Integration ◽  
Voltage Limit ◽  
The Stability ◽  
Scale Integration

Renewable energy systems (RESs), such as photovoltaic (PV) systems, are providing increasingly larger shares of power generation. PV systems are the fastest growing generation technology today with almost ~30% increase since 2015 reaching 509.3 GWp worldwide capacity by the end of 2018 and predicted to reach 1000 GWp by 2022. Due to the fluctuating and intermittent nature of PV systems, their large-scale integration into the grid poses momentous challenges. This paper provides a review of the technical challenges, such as frequency disturbances and voltage limit violation, related to the stability issues due to the large-scale and intensive PV system penetration into the power network. Possible solutions that mitigate the effect of large-scale PV system integration on the grid are also reviewed. Finally, power system stability when faults occur are outlined as well as their respective achievable solutions.

scholarly journals Forecasting Day-Ahead Hourly Photovoltaic Power Generation Using Convolutional Self-Attention Based Long Short-Term Memory

Energies ◽  
2020 ◽  
Vol 13 (15) ◽  
pp. 4017 ◽  
Author(s):  
Dukhwan Yu ◽  
Wonik Choi ◽  
Myoungsoo Kim ◽  
Ling Liu
Keyword(s):  
Power Generation ◽  
Deep Learning ◽  
Prediction Accuracy ◽  
Short Term Memory ◽  
Local Context ◽  
Short Term ◽  
Term Memory ◽  
Proposed Model ◽  
Long Short Term Memory ◽  
Pv Power Generation

The problem of Photovoltaic (PV) power generation forecasting is becoming crucial as the penetration level of Distributed Energy Resources (DERs) increases in microgrids and Virtual Power Plants (VPPs). In order to improve the stability of power systems, a fair amount of research has been proposed for increasing prediction performance in practical environments through statistical, machine learning, deep learning, and hybrid approaches. Despite these efforts, the problem of forecasting PV power generation remains to be challenging in power system operations since existing methods show limited accuracy and thus are not sufficiently practical enough to be widely deployed. Many existing methods using long historical data suffer from the long-term dependency problem and are not able to produce high prediction accuracy due to their failure to fully utilize all features of long sequence inputs. To address this problem, we propose a deep learning-based PV power generation forecasting model called Convolutional Self-Attention based Long Short-Term Memory (LSTM). By using the convolutional self-attention mechanism, we can significantly improve prediction accuracy by capturing the local context of the data and generating keys and queries that fit the local context. To validate the applicability of the proposed model, we conduct extensive experiments on both PV power generation forecasting using a real world dataset and power consumption forecasting. The experimental results of power generation forecasting using the real world datasets show that the MAPEs of the proposed model are much lower, in fact by 7.7%, 6%, 3.9% compared to the Deep Neural Network (DNN), LSTM and LSTM with the canonical self-attention, respectively. As for power consumption forecasting, the proposed model exhibits 32%, 17% and 44% lower Mean Absolute Percentage Error (MAPE) than the DNN, LSTM and LSTM with the canonical self-attention, respectively.

scholarly journals The Technical Challenges Facing the Integration of Small-Scale and Large-scale PV Systems into the Grid: A Critical Review

Electronics ◽  
2019 ◽  
Vol 8 (12) ◽  
pp. 1443 ◽  
Author(s):  
Abdullah Alshahrani ◽  
Siddig Omer ◽  
Yuehong Su ◽  
Elamin Mohamed ◽  
Saleh Alotaibi
Keyword(s):  
Large Scale ◽  
Distribution Network ◽  
Small Scale ◽  
Energy Access ◽  
Solar Pv ◽  
Pv System ◽  
Pv Systems ◽  
Technical Challenges ◽  
Technical Solutions ◽  
The Impact

Decarbonisation, energy security and expanding energy access are the main driving forces behind the worldwide increasing attention in renewable energy. This paper focuses on the solar photovoltaic (PV) technology because, currently, it has the most attention in the energy sector due to the sharp drop in the solar PV system cost, which was one of the main barriers of PV large-scale deployment. Firstly, this paper extensively reviews the technical challenges, potential technical solutions and the research carried out in integrating high shares of small-scale PV systems into the distribution network of the grid in order to give a clearer picture of the impact since most of the PV systems installations were at small scales and connected into the distribution network. The paper reviews the localised technical challenges, grid stability challenges and technical solutions on integrating large-scale PV systems into the transmission network of the grid. In addition, the current practices for managing the variability of large-scale PV systems by the grid operators are discussed. Finally, this paper concludes by summarising the critical technical aspects facing the integration of the PV system depending on their size into the grid, in which it provides a strong point of reference and a useful framework for the researchers planning to exploit this field further on.

Sign in / Sign up

Export Citation Format

Share Document