Indonesian Journal of Electrical Engineering and Computer Science Vol. 23, No. 1, July 2021, pp. 110~119 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v23.i1.pp110-119  110 Journal homepage: http://ijeecs.iaescore.com Management switching angles real-time prediction by artificial neural network Mohammed Rasheed Jubair Al-Hiealy, Mohammad Shahir Bin Abdul Majed Shikh, Abdurrahman Bin Jalil, Suhaila Abdul Rahman, Muath Jarrah Departmnet of Computer Engineering, University Malaysia of Computer Science & Engineering (UNIMY), Selangor, Malaysia Article Info ABSTRACT Article history: Received Nov 7, 2020 Revised Apr 27, 2021 Accepted May 10, 2021 Artificial neural networks (ANNs) is an efficient way for different types of real-world prediction problems. In the past decade, it has given a tremendous surge in a global research activities. ANNs embody much certainty and provide a great deal of promise This paper has present artificial neural network (ANN) technique analysis and prediction for management switching angles real-time. The proposes to be used ANN for prediction and selected obtine angles for implement the timing diagram for mulitlvel inverter circuit. In order to control the fundamental component, ANNs are used to solve the analysis of non-linear equation of the output timing diagram in order to determine the switching angles. Substantially, the number of switching devices are reducing as possible basically for reducing a switching loss in the system, also have been used ANNs technique to optimize a switching angles behavior to reduce total harmonic distortion (THD) at voltage and current output waveform equal THDV 8.05% THDA 5.1%. For the proposed controllers, the performance and results by the ANNs were obtained and compared by using MATLAB software. Keywords: Artificial intelligence Harmonics optimization Neural network ANN Switching angle This is an open access article under the CC BY-SA license. Corresponding Author: Mohammed Rasheed Jubair Al-Hiealy Department of Computer Engineering University Malaysia of Computer Science & Engineering (UNIMY) Selangor, Malaysia Email: rasheed.alhiealy@unimy.edu.my 1. INTRODUCTION A multilevel inverter is an electronic circuit with multi-level power, this inverter/circuit can be converted into a direct power (DC) output supply and supplies an alternate voltage (AC) waveform. ”Renewable sources of energy such as photovoltaics, wind and water turbines are now essential energy sources due to their renewable and environmental friendliness” [1]-[3]. Most of these renewable pathways generate a DC waveform as an output that is not appropriate for transmissions, distributions and uses. Based on this, the DC output power should be converted to AC power waveform and this can be achieved through the MLI circuit [2]-[5]. In general, based on their principal structure, MLI structures have been categorized into three main topologies namely cascaded h”bridge (CHB-MLI) which is the most common and reliable, diode-clamped (DC-MLI) and fly”capacitor (FC-MLI). A connected cells in series in the load compose a CHB-MLI topology, while conventional structure of CHB-MLI uses four switches producing three levels for each cell [4], [5]. Several modulation methods for eliminating a harmonic during the output waveform, such as the pulse wide module (PWM), called a high frequency switching technique, including S-PWM and SV- PWM have been used. Secondly, selective harmonic elimination (SHE-PWM) method, often called the best