Novel Application of Relief Algorithm in Cascade ANN Model for Prognosis of Photovoltaic Maximum Power Under Sunny Outdoor Condition of Sikkim India: A Case Study

The article presents calculations and power flow of a real virtual power plant (VPP), containing a fragment of low and medium voltage distribution network. The VPP contains a hydropower plant (HPP), a photovoltaic system (PV) and energy storage system (ESS). The purpose of this article is to summarize the requirements for connection of generating units to the grid. Paper discusses the impact of the requirements on the maximum installed capacity of distributed energy resource (DER) systems and on the parameters of the energy storage unit. Firstly, a comprehensive review of VPP definitions, aims, as well as the characteristics of the investigated case study of the VPP project is presented. Then, requirements related to the regulation, protection and integration of DER and ESS with power systems are discussed. Finally, investigations related to influence of DER and ESS on power network condition are presented. One of the outcomes of the paper is the method of identifying the maximum power capacity of DER and ESS in accordance with technical network requirements. The applied method uses analytic calculations, as well as simulations using Matlab environment, combined with real measurement data. The obtained results allow the influence of the operating conditions of particular DER and ESS on power flow and voltage condition to be identified, the maximum power capacity of ESS intended for the planed VPP to be determined, as well as the influence of power control strategies implemented in a PV power plant on resources available for the planning and control of a VPP to be specified. Technical limitations of the DER and ESS are used as input conditions for the economic simulations presented in the accompanying paper, which is focused on investigations of economic efficiency.

Download Full-text

A GA trained ANN model for maximum power point tracking in solar photo voltaic systems

International Journal of Engineering & Technology ◽

10.14419/ijet.v7i1.9.10383 ◽

2018 ◽

Vol 7 (1.9) ◽

pp. 294 ◽

Cited By ~ 1

Author(s):

V. Dega Rajaji ◽

K. Chandra Sekhar

Keyword(s):

Maximum Power Point Tracking ◽

Maximum Power ◽

Reference Voltage ◽

Ann Model ◽

Maximum Power Point ◽

Data Set ◽

The Real ◽

Point Tracking ◽

Power Point Tracking ◽

Power Point

Sunlight based vitality is one of the imperative segment of sustainable power source assets and reaping of sun oriented vitality is an actually difficult assignment. Sunlight based PV cells display non liner conduct and their execution is affected by an assortment of components. Fluctuating insolation and temperature assumes a critical part in characterizing the working purpose of the PV cell. Most extreme Power Point Tracking (MPPT) calculations are important to amplify the yield control. The target of getting MPP in PV frameworks is to manage the real working voltage of PV boards. The fundamental reason for acquiring MPP in PV frameworks is to direct the real working voltage of PV boards. In this paper a Genetic Algorithm (GA) –Artificial Neural Network (ANN) MPPT approach is presented. GA is employed to adaptively change the weights during the course of ANN training. The proposed approach is validated by comparing its performance against a data set which contains voltages forecast by ANN approach, Adaptive Neuro Fuzzy (ANFIS) approach and reference voltage obtained through mathematical modeling. The results demonstrate the suitability of the proposed approach in maximum power point tracking through its better performance when compared to the other two approaches. Pearson product-moment correlation has also been calculated to compare the correlation between the predicted voltage and the reference voltage.

Download Full-text