ThesisPDF Available

Towards Efficient Energy Management: Prediction Based Generation and Secure Trading

Authors:

Abstract

With the advent of the smart grid (SG), the concept of energy management flourished rapidly and it gained the attention of researchers. Forecasting plays an important role in energy management. In this work, a recurrent neural network, long short term memory (LSTM), is used for electricity price and demand forecasting using big data. This model uses multiple variables as input and forecasts the future values of electricity demand and price. Its hyperparameters are tuned using the Jaya optimization algorithm to improve the forecasting ability. It is named as Jaya LSTM (JLSTM). Moreover, the concept of local energy generation using renewable energy sources is also getting popular. In this work, to implement a hybrid peer to peer energy trading market, a blockchain based system is proposed. It is fully decentralized and allows the market members to interact with each other and trade energy without involving a third party. In addition, in vehicle to grid and vehicle to vehicle energy trading environments, local aggregators perform the role of energy brokers and are responsible for validating the energy trading requests. A solution to find accurate distance with required expenses and time to reach the charging destination is also proposed, which effectively guides electric vehicles (EVs) to reach the relevant charging station and encourages energy trading. Moreover, a fair payment mechanism using a smart contract to avoid financial irregularities is proposed. Apart from this, a blockchain based trust management method for agents in a multi-agent system is proposed. In this system, three objectives are achieved: trust, cooperation and privacy. The trust of agents depends on the credibility of trust evaluators, which is verified using the proposed methods of trust distortion, consistency and reliability. To enhance the cooperation between agents, a tit-3-for-tat repeated game strategy is developed. The strategy is more forgiving than the existing tit-for-tat strategy. It encourages cheating agents to re-establish their trust by cooperating for three consecutive rounds of play. Also, a proof-of-cooperation consensus protocol is proposed to improve agents’ cooperation while creating and validating blocks. The privacy of agents is preserved in this work using the publicly verifiable secret sharing mechanism. Additionally, a blockchain based edge and cloud system is proposed to resolve the resource management problem of EVs in a vehicular energy network. Firstly, a min-max optimization problem is formulated to construct the proposed entropy based fairness metric for resource allocation. This metric is used to determine whether users have received a fair share of the system’s resources or not. Secondly, a new deep reinforcement learning based content caching and computation offloading approach is designed for resource management of EVs. Lastly, a proof-of-bargaining consensus mechanism is designed for block’s validation and selection of miners using the concept of iterative negotiation. Besides, a survey of electricity load and price forecasting models is presented. The focus of this survey is on the optimization methods, which are used to tune the hyperparameters of the forecasting models. Moreover, this work provides a systematic literature review of scalability issues of the blockchain by scrutinizing across multiple domains and discusses their solutions. Finally, future research directions for both topics are discussed in detail. To prove the effectiveness of the proposed energy management solutions, simulation are performed. The simulation results show that the energy is efficiently managed while ensuring secure trading between energy prosumers and fair resource allocation.
A preview of the PDF is not available
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The real time Vehicle-to-Grid (V2G) and Vehicle-to-Vehicle (V2V) energy trading have ameliorated the smart city environment. Electric Vehicles (EVs) in smart cities have made it possible to balance the energy demand and supply without burdening the power grids. They also help in minimizing both the heap amounts of pollution and the greenhouse gas emissions. EVs not only charge their batteries from Charging Stations (CSs), but when the smart city faces energy deficiency, they supply their surplus energy to the power grids to fulfill energy demand as well. In addition, EVs also balance the energy demand and supply through energy trading at the local level during peak hours, which strengthens V2V energy trading. However, some issues, such as security threats and privacy leakage prevent EVs from participating in the energy trading process. In addition, the lack of incentives and knowledge about the cost of reaching the energy trading place within minimum time are also big challenges. Therefore , to solve the aforementioned challenges, a secure and efficient scheme for V2V and V2G energy trading is proposed in this paper. The proposed scheme also helps in promoting environmental friendliness and making the smart cities sustainable and reliable. In the proposed scheme, energy trading transactions are secured using consortium blockchain, wherein the Local Aggregators (LAGs) are selected as authorized nodes. LAGs perform their role as energy brokers and are responsible for validating the energy trading requests using Proof of Authority (PoA) consensus mechanism. Moreover, a solution to find accurate distance with required expenses and time to reach the charging destination is also proposed , which effectively guides EVs to reach the relevant CSs and encourages energy trading. Besides, we propose a fair payment mechanism using a smart contract to avoid financial irregularities. An incentive provisioning mechanism is also given in the proposed work to prevent EVs from acting selfishly. The efficient power flow having minimized losses in the vehicular network is of much importance. Therefore, the energy losses incurred in both V2G and V2V are discussed in this work. Furthermore, Oyente is used for smart contract's security analysis and for testing Ethereum's resilience against different security flaws. Two attacker models are proposed and the security analyses of the models are * Corresponding author also provided. The analyses show that the proposed system is robust against the double spending and Sybil attacks. Finally, the efficient performance of our proposed scheme is validated and analyzed. The simulation results proved that our proposed work outperforms existing work in terms of providing a secure and efficient energy trading platform for both V2G and V2V environments.
Article
Full-text available
Vehicular networks as the key enablers in Intelligent Transportation Systems (ITS) and the Internet of Things (IoT) are key components of smart sustainable cities. Vehicles as a significant component of smart cities have emerging in-vehicle applications that can assist in good governance for sustainable smart cities. Most of these applications are delay sensitive and demand high computational capabilities that are provided by emerging technologies. Utilizing the distributed computational resources of vehicles with the help of volunteer computing is an efficient method to fulfill the high computational requirements of vehicles itself and the other components of smart cities. Vehicle as a resource is an emerging concept that must be considered to address the future challenges of sustainable smart cities. In this paper, an infrastructure-assisted job scheduling and task coordination mechanism in volunteer computing-based VANET called RSU-based VCBV is proposed, which enhances the architecture of VANET to utilize the surplus resources of vehicles for task execution. We propose job scheduling and task coordination algorithms for different volunteer models. Further, we design and implement an adaptive task replication method to seek fault tolerance by avoiding task failures due to locations of vehicles. We propose a task replication algorithm called location-based task replication algorithm. Extensive simulations validate the performance of our proposed volunteer models while comparing average task execution time and weight ratios with existing work.
Article
Full-text available
In this paper, we propose a mobile edge computing (MEC)-enabled unmanned aerial vehicle (UAV)-assisted vehicular ad hoc network (VANET) architecture, based on which a number of vehicles are served by UAVs equipped with computation resource. Each vehicle has to offload its computing tasks to the proper MEC server on the UAV due to the limited computation ability. To counter the problems above, we first model and analyze the transmission model and the security assurance model from the vehicle to the MEC server on UAV, and the task computation model of the local vehicle and the edge UAV. Then, the vehicle offloading problem is formulated as a multi-objective optimization problem by jointly considering the task offloading, the resource allocation, and the security assurance. For tackling this hard problem, we decouple the multi-objective optimization problem as two subproblems and propose an efficient iterative algorithm to jointly make the MEC selection decision based on the criteria of load balancing and optimize the offloading ratio and the computation resource according to the Lagrangian dual decomposition. Finally, the simulation results demonstrate that our proposed scheme achieves significant performance superiority compared with other schemes in terms of the successful task processing ratio and the task processing delay.
Article
Full-text available
The drastic increase in real-time vehicle generated data of various types has imparted a great concept of data trading in vehicular networks. Whereas immense usage of Electric Vehicles (EVs) as mobile energy carriers have supported distributed energy trading due to their bidirectional charging and discharging capabilities. The trustless environment of Internet of Electric Vehicles (IoEV), including fuel vehicles and EVs, encounters trading disputes and conflicting interests among trading parties. To address these challenges, we exploit consortium blockchain to maintain transparency and trust in trading activities. Smart contracts are used to tackle trading disputes and illegal actions. Data duplication problem occurs when a dishonest user sell previously traded data multiple times for financial gain. Therefore, data duplication validation is done through previously stored hash-list at roadside units (RSUs) employed with bloom filters for efficient data lookup. Removing data duplication at an earlier stage reduces storage cost. Moreover, an elliptic curve bilinear pairing based digital signature scheme is used to ensure the reliability and integrity of traded data. To ensure persistent availability of traded data, InterPlanetary File System (IPFS) is used, which provides fault-tolerant and a reliable data storage without any single point of failure. On the other hand, the energy trading transactions among EVs face some security and privacy protection challenges. An adversary can infer the energy trading records of EVs, and launch the data linkage attacks. To address this issue, an account generation technique is used that hides the energy trading trends. The new account generation for an EV depends upon its traded volume of energy. The experimental results verify the efficiency of the proposed data and energy trading scheme in IoEV with the reliable and secure data storage.
Article
Full-text available
In Vehicular Ad-hoc Networks (VANETs), a large amount of data is shared between vehicles and Road Side Units (RSUs) in real-time. VANETs assist in sharing traffic information effectively and timely to improve traffic efficiency and reliability. However, less storage capability and selfish behavior of the vehicles are important issues that need to be tackled. The traditional storage mechanisms involve third parties for data management, which are insecure, untrustworthy, non-transparent, and unreliable. To overcome these issues, a blockchain-based data storage scheme for VANETs is proposed in this paper. It exploits the benefits of the Interplanetary File System (IPFS) and blockchain is implemented on RSUs. The RSUs receive the aggregation packets sent by vehicles. These packets contain the events' information that occur in vehicles' surroundings. After verifying an aggregation packet, the RSUs store the event's information in IPFS and the reputation values of the sender vehicle in the blockchain. The reputation value is calculated based on the witnesses' (others vehicles) opinion, whether they agree with the initiator or not about an event. The initiator is the vehicle who initializes the event. Moreover, an incentive mechanism is also proposed in this work in which monetary incentives are given to the repliers who respond to the event information. These incentives are given by the initia-tors after verifying the signatures of the repliers. All the transactions involved in the incentive process are stored in the blockchain. Finally, Oyente is used for the security analysis of the proposed smart contracts. A comparison of the proposed scheme with the logistic regression scheme is also presented.
Article
Full-text available
The significance of electricity cannot be overlooked as all fields of life like material production, health care, educational sector, etc., depend upon it to render consistent and high-quality services, increase productivity and business continuity. To this end, energy operators have experienced a continuous increasing trend in the electricity demand for the past few decades. This may cause many issues like load shedding, increased electricity bills, imbalance between supply and demand, etc. Therefore, forecasting of electricity demand using efficient techniques is crucial for the energy operators to decide about optimal unit commitment and to make electricity dispatch plans. It also helps to avoid wastage as well as the shortage of energy. In this study, a novel forecasting model, known as ELS-net is proposed, which is a combination of an Ensemble Empirical Mode Decomposition (EEMD) method, multi-model Ensemble Bi Long Short-Term Memory (EBiLSTM) forecasting technique and Support Vector Machine (SVM). In the proposed model, EEMD is used to distinguish between linear and non-linear intrinsic mode functions (IMFs), EBiLSTM is used to forecast the non-linear IMFs and SVM is employed to forecast the linear IMFs. Using separate forecasting techniques for linear and non-linear IMFs decreases the computational complexity of the model. Moreover, SVM requires low computational time as compared to EBiLSTM for linear IMFs. Simulations are performed to examine the effectiveness of the proposed model using two different datasets: New South Wales (NSW) and Victoria (VIC). For performance evaluation, Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) are used as performance metrics. From the simulation results, it is obvious that the proposed ELS-net model outperforms the start-of-the-art techniques, such as EMD-BILSTM-SVM, EMD-PSO-GA-SVR, BiLSTM, MLP and SVM in terms of forecasting accuracy and minimum execution time.
Article
Full-text available
Deep neural networks excel at finding hierarchical representations that solve complex tasks over large datasets. How can we humans understand these learned representations? In this work, we present network dissection, an analytic framework to systematically identify the semantics of individual hidden units within image classification and image generation networks. First, we analyze a convolutional neural network (CNN) trained on scene classification and discover units that match a diverse set of object concepts. We find evidence that the network has learned many object classes that play crucial roles in classifying scene classes. Second, we use a similar analytic method to analyze a generative adversarial network (GAN) model trained to generate scenes. By analyzing changes made when small sets of units are activated or deactivated, we find that objects can be added and removed from the output scenes while adapting to the context. Finally, we apply our analytic framework to understanding adversarial attacks and to semantic image editing.
Article
Full-text available
In recent years, the non-handcrafted feature extraction methods have gained increasing popularity for solving pattern classification tasks due to their inherent ability to extract robust features and handle outliers. However, the design of such features demands a large set of training data. Meta-heuristic optimization schemes can facilitate feature learning even with a small amount of training data. This paper presents a new feature learning mechanism called multi-objective Jaya convolutional network (MJCN) that attempts to learn meaningful features directly from the images. The proposed scheme, unlike the convolutional neural networks, comprises a convolution layer, a multiplication layer, an activation layer and an optimizer known as multi-objective Jaya optimizer (MJO). The convolution layer searches meaningful patterns in an image through the local neighborhood connections and the multiplication layer projects the convolutional response to a more compact feature space. The weights used in these layers are initialized randomly and MJO is then introduced to optimize the weights. The main objective of MJO is to maximize the inter-class distance and minimize the intra-class variance. The feature vectors are finally derived using the optimized weights. The derived features are finally fed to a set of standard classifiers for recognition of characters. The performance of the proposed model is evaluated on various benchmark datasets, namely, NITR Odia handwritten character, ISI Kolkata Odia numeral, ISI Kolkata Bangla numeral, and MNIST as well as a newly developed dataset NITR Bangla numeral. The experimental results show that the proposed scheme outperforms other state-of-the-art approaches in terms of recognition accuracy.
Conference Paper
Prosumer consortium energy transactive models can be one of the solutions for energy costs, increasing performance and for providing reliable electricity utilizing distributed power generation, to a local group or community, like a university. This research study demonstrates the simulation of blockchain based power trading, supplemented by the solar power prediction using MLFF neural network training in two prosumer nodes. This study can be the initial step in the implementation of a power trading market model based on a decentralized blockchain system, with distributed generations in a university grid system. This system can balance the electricity demand and supply within the institute, enable secure and rapid transactions, and the local market system can be reinforced by forecasting solar generation. The performance of the MLFF training can predict almost 90% accuracy of the model as short term ahead forecasting. Because of it, the prosumer bodies can complete the decision making before trading to their benefit.
Article
Renewable energy resources are natural, clean, economical, and never-ending energy resources. Wind energy is an important clean, cheap, and easy applicable energy sources. On account of this, generation of the energy from wind technology has been raised day by day because of the competition with fossil-fuel power production methods. By depending on increases the number of turbines located in the wind farm, the average power obtains from each wind turbine appreciable reduces due to the existence of wake effects within the wind farm. Therefore, the optimal placement of turbines in a wind farm provides to get optimum wind energy from the wind farm. When the place where the wind turbines are located is considered as NxN grid, a wind turbine can be established to each cell of this grid. Whether a wind turbine is replaced to each cell of the grid or not can be modeled as a binary-based optimization problem. In this study, a Jaya-based binary optimization algorithm is proposed to determine which cells are used for wind turbine replacement. In order to justify the efficiency of the proposed approach, two different test cases are considered, and the solutions produced by the proposed approach are compared with the solutions of the swarm intelligence or evolutionary computation methods. According to the experiments and comparisons the Jaya-based binary approach shows a superior performance than compared approaches in terms of cost and power effectiveness. While the efficiency of the Jaya-based approach is 92.2% with 30 turbines replacement on 10 × 10 grid, the efficiency of the Jaya-based binary method is 95.7% with 43 turbines replacement on 20 × 20 grid.