Study and analysis of SARIMA and LSTM in forecasting time series data

As the technology is advancing in each and every field, it also advanced in making houses smarter where one can control the appliance remotely. Here the word smart also helps in generating the electricity bill in a way of reducing the human effort in calculating all the necessary parameters. But not every person knows how electricity works and how the bill is generated. So, the main objective of this project is to let every user know how the consumption of each appliance effect the electricity bill. This project also helps the user to control the home appliance remotely using a smart phone. Accordingly, this research work has developed a Smart system for leveraging a controlled and effective billing. Generally, disabled or elderly people find it difficult to near the switch and turn it on/off in a home. So, the application designed in this paper will fulfill the need. Current and Voltage sensors are used to get the values of voltage and current which will be further used in the generation of bill. The values from the sensors are communicated to the Arduino UNO through I/O ports and the bill is generated. Here, a Bluetooth module is used which acts as a medium and helps the data gets transmitted. All the necessary controls of the appliance and its energy consumption will be present in an Android Application.

Arduino based Smart System for Control and Effective Billing

Cost/comfort-oriented clustering-based extended time of use pricing

The demand for electricity power consumption (PC) is growing lots due to the upward thrust of hardware and the development of the population. Hence it is mandatory to predict energy to enhance its management and co-operation among the energy utilized in a construction and the power grid. There is a problem for predicting the Energy Consumption Prediction (ECP) due to tremendous problems like climate conditions and the effective behaviour of tenants. We present a keen crossover approach that joins Convolutional Neural Network (CNN) with Gated Recurrent Unit (GRU) method which is a Deep Learning technique through three steps which achieves better prediction results when compared to other existing techniques.

Household Electricity Power Consumption Prediction Using CNN-GRU Techniques

A newly emerging technology which provides resources for the purpose of computation of resources which is based on the user demand is known as Cloud computing. This supply based on the demand for resources are scheduled on the basis of certain policies. In order to combat and solve the Virtual Machine (VM) scheduling problem in cloud computing, a new and unique Tabu Search (TS) resource based scheduling algorithm which has its basis on auction mechanism is shown by keeping in mind many factors including the band with and deadline for network and auction. The bids of the clients are firstly sequenced within the competition deadline. Secondly, a screening of the client group is done and the VM resource that corresponds to it is configured in accordance to the minimal cost of the service providers. Lastly, payment price is finalised by taking average payment as well as payment made by competitors into consideration thereby enabling the timely completion of the VM resource task. The results of the simulation experiments prove that the algorithm proposal that was made is capable of enhancing efficiently the QoS or the Quality of Service of the cloud environment, and also ensure that the profits of the providers of cloud service and the rate of resource utilization of VM are efficiently completed.

Optimised scheduling in cloud computing

In Smart Grid, the Smart meters assume different role with smart capacities so as to satisfy the consumer’s needs. The Smart meter can gauge and impart detailed real time power utilization, empower remote continuous checking and control utilizations. The consumers are given ongoing valuing and analyses devoured data, with its technical information which is transmitted from the network, and furthermore give progressively itemized input on every machine to the consumer. This paper gives a review of the prediction and monitoring with respect to control utilization. It is focused to real time scenario like Smart metering and home apparatus like home sight and sound substance over the power utilization grid.

Smart Electricity meter on Real time Price Forecasting and Monitoring system

Cloud internet applications have recently attracted a large number of users in the Internet. With the invention of these cloud internet applications, it is inefﬁcient to allocate maximum number of participants in real time group auction system. So an efficient approximation algorithm is proposed with the improved combinatorial double auction protocol. It is developed to enable different kinds of resource distribution among multiple users and providers. At the same time it includes more number of participants in an auction. Due to the NP-hardness of binary integer programming for resource distribution in a real time group auction system, the improved approximation algorithm is proposed to deal with np-hardness and to obtain the optimal solution. Participant honesty is necessary to ensure auction trustfulness.

/pdf/enhanced-real-time-group-auction-system-for-efficient-2fbjuuzsli.pdf

V. Sakthi Aravind

Papers

Smart Electricity meter on Real time Price Forecasting and Monitoring system

Enhanced Real-Time Group Auction System for Efficient Allocation of Cloud Internet Applications