Selection and development of neural network architecture for finding the maximum mechanical power of the propeller.

Abstract

An example of using a neural network for finding the value of the point of maximum power in a power supply system with an alternative energy source (wind generator) is given in this work. The main feature of the built neural network is its structure. Neural network-multilayer (the first two layers have the activation function - hyperbolic tanges, and the third - a linear function). For the successful integration of the network into the system was considered the process of learning NM, construction of it is carried out in the application package Matlab. In this paper, a neural network was developed and modeled to find the maximum value of the mechanical power of the wind wheel. This network is multi-layered; weight coefficients and bias obtained by experiment were used for construction. This network takes into account the change in coefficients, which means it can be retrained. This property is very useful when an error occurs in the real equipped, caused by the aging of the equipment.

Authors and Affiliations

Dmytro Ostrenko, Oleksandr Kollarov

Keywords

Related Articles

Generalized dependences of turbogenerator electromagnetic parameters on saturation.

For the first time we obtained generalized analytical dependences of the parameters of equivalent circuits (which take into account different mutual inductive coupling between the windings) of TBB turbogenerators...

Influence on the durability of work of high-temperature rotors of long-term static strength and low cycle fatigue

A significant part of heat and mechanical equipment at Ukrainian thermal power plants has reached the end of normative and extended operating life. The power plants were commissioned 30-40 years ago and now appear to be...

The converter unit with function of voltage stabilization for a combined power supply system with renewable power sources.

Combined power supply systems with renewable energy sources are used for local facilities in order to ensure stable power supply by using existing overloaded networks remote from transformer substations that are charac-...

A method of industrial oriented software influence evaluation in manufacturing enterprises information systems.

A method of calculating the volume of data transferred through industrial data transfer network during working of SCADA/HMI-, EAM-, MES-systems and OPC-servers in industrial and office segments...

Proposals on the prevention and elimination of defects in the cores of turbogenerators stators.

Condition of domestic turbogenerators demands the constant control of parameters, technical maintenance and rehabilitation. The works to assess the status of turbogenerators is carried out using statistica...

Download PDF file
  • EP ID EP660518
  • DOI -
  • Views 181
  • Downloads 0

How To Cite

Dmytro Ostrenko, Oleksandr Kollarov (2018). Selection and development of neural network architecture for finding the maximum mechanical power of the propeller.. Наукові праці Донецького Національного Технічного Університету серія: Електротехніка і Енергетика, 1(1), 63-67. https://europub.co.uk/articles/-A-660518