Development of intelligent controller for high performance electric drives with hybrid CSA and RERNN technique
Abstract Electric drives play a crucial role in various industrial applications, requiring precise control and adaptability to changing parameters. This research proposes a hybrid approach combining Circle Search Algorithm (CSA) and Recalling-Enhanced Recurrent Neural Network (RERNN), termed CSA-RER...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-04-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-98704-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Electric drives play a crucial role in various industrial applications, requiring precise control and adaptability to changing parameters. This research proposes a hybrid approach combining Circle Search Algorithm (CSA) and Recalling-Enhanced Recurrent Neural Network (RERNN), termed CSA-RERNN, for intelligent controller-fed electric drives. The primary objective is to control the speed of electric drives using complex mechanical configurations and variable parameters. The CSA optimizes the voltage source inverter’s control signal, while RERNN predicts the optimal control signal. The system considered the variables of inertia, load torque, and rotor shaft angle. According to MATLAB/Simulink, the proposed method demonstrates superior performance compared to existing techniques, such as Singular Spectrum Analysis (SSA), Artificial Neural Network (ANN), and Bee Colony Optimization (BCO). The experimental results demonstrate that the CSA-RERNN approach achieves lower error rates and better speed control. The system maintains response characteristics despite changes in the moment of inertia of the object, thereby ensuring long-term performance stability without deactivating the adaptation mechanism. From the experiment it is evident that the system’s performance remains stable despite fluctuations in the moment of inertia, which changes between 1.2 kgm2 and 2.3 kgm2 over time. In addition, the system’s response characteristics, including speed and current waveforms, are consistent and stable. At 1976s, the reference angular speed (ωRef) reaches 0.9 rad/s, and the system adjusts accordingly. At the same time, the current varies, with the highest reading reaching 6 A and the lowest dipping to −6 A. Additionally, the proposed method proves more cost-effective than conventional intelligent controllers. Overall, the CSA-RERNN technique offers a robust and efficient solution for electric drive control in various industrial applications. |
|---|---|
| ISSN: | 2045-2322 |