TL;DR
This paper introduces a novel multi-objective control strategy for Switched Reluctance Generators (SRGs) in small-scale wind power systems. By optimizing both turn-off and turn-on angles using a Simulated Annealing Algorithm (SAA) and a carefully crafted multi-objective function, the strategy simultaneously improves system efficiency, reduces power converter losses, and maintains stable output power, even with fluctuating wind speeds.
What Problem Does This Paper Address?
Switched Reluctance Generators (SRGs) are attractive for wind power generation due to their simple structure and robustness. However, achieving optimal performance in SRGs is challenging due to inherent trade-offs. Specifically, maximizing system efficiency and minimizing output voltage ripple are often conflicting goals. This means that traditional single-objective optimization methods can lead to compromises in other critical performance metrics. The paper aims to address this by developing a control strategy that balances these competing objectives for improved overall SRG performance in small-scale wind power applications.
Methods
The proposed control strategy employs a two-stage optimization approach using the Simulated Annealing Algorithm (SAA):
- Off-line Turn-Off Angle Optimization: The turn-off angle is optimized offline for each rotor speed to maximize the output power. The results are then fitted into a function of rotor speed for real-time online regulation. This step ensures the generator operates within its optimal power generation range.
- Online Multi-Objective Optimization of Turn-On Angle: A closed-loop controller regulates the reference current to meet the required output power. Simultaneously, a multi-objective function is defined, considering system efficiency, output voltage ripple, and power converter loss. The turn-on angle is then tuned online by SAA based on this multi-objective evaluation, aiming to find the best compromise between these performance indicators.
The multi-objective function is defined as: $f(x) = \lambda_1 \frac{\eta}{(\eta)_{max}} + \lambda_2 \frac{1/(\Delta U_{dc})_{max}}{1/\Delta U_{dc}} + \lambda_3 \frac{1/(APPC)_{max}}{1/APPC}$ where $\lambda_1, \lambda_2, \lambda_3$ are weight coefficients for system efficiency, voltage ripple, and power converter loss, respectively. The paper sets these as $\lambda_1 = 0.5$, $\lambda_2 = 0.3$, and $\lambda_3 = 0.2$.
Datasets
Not discussed in the paper. The study relies on simulation and experimental verification using a specific SRG platform.
Key Insights and Ideas
- The inherent trade-off: The paper clearly illustrates the fundamental conflict between maximizing efficiency and minimizing voltage ripple in SRGs, a key challenge for their high-performance operation.
- SAA for robust optimization: Using SAA for both off-line and on-line optimization is crucial. It helps avoid local optima and allows for dynamic adjustments in response to changing operating conditions.
- Multi-objective function is key: The development of a weighted multi-objective function that balances efficiency, voltage ripple, and power converter loss is the core innovation. This allows for a more holistic approach to performance improvement than single-objective methods.
- Online adaptation is vital: The on-line tuning of the turn-on angle ensures the SRG can adapt to dynamic changes in wind speed and load, maintaining stable output power.
- Voltage ripple as a sacrifice: The results suggest that to achieve better overall performance (efficiency and reduced loss), a slight increase in voltage ripple might be an acceptable trade-off, which is a pragmatic approach for real-world applications.
Key Results
The simulation and experimental results demonstrate the effectiveness of the proposed multi-objective control strategy:
- Improved System Efficiency: The strategy leads to a noticeable improvement in system efficiency across various operating speeds and output powers, both before and after optimization. For instance, at 800 rpm and 50W output, efficiency is improved after optimization.
- Reduced Power Converter Loss: Power converter losses are consistently reduced after applying the multi-objective control. For example, at 1000 rpm and 100W output, power converter loss decreases significantly.
- Stable Output Power: The controller effectively maintains the output power at the target value, even during dynamic changes in wind speed and load. The maximum output power variation range is set to 5W, and re-optimization is triggered if disturbances exceed this.
- Voltage Ripple Trade-off: While system efficiency and power converter loss improve, the output voltage ripple generally increases. However, this increase is considered acceptable in exchange for the overall performance gains.
Limitations and Future Work
- Voltage Ripple Increase: The paper acknowledges that the output voltage ripple increases with the proposed strategy. While this is a controlled trade-off for overall performance, it might be a limitation for applications with extremely strict voltage ripple requirements.
- Complexity of SAA: While SAA is effective, its computational demands for real-time optimization might be a consideration for very low-cost or resource-constrained embedded systems.
- Future Work: The authors suggest that the proposed solution can be flexibly applied to SRGs with different structures, indicating potential for broader adoption. Further research could explore more advanced optimization algorithms or adaptive weighting schemes for the multi-objective function to further refine performance. Investigating the impact of different weight coefficient distributions on specific applications could also be a valuable direction.
Summary updated