Data-driven Bi-level Optimization of Thermal Power Systems with embedded Artificial Neural Networks
arXiv:2602.13746v1 Announce Type: new Abstract: Industrial thermal power systems have coupled performance variables with hierarchical order of importance, making their simultaneous optimization computationally challenging or infeasible. This barrier limits the integrated and computationally scaleable operation optimization of industrial thermal power systems. To address this issue for large-scale engineering systems, we present a fully machine learning-powered bi-level optimization framework for data-driven optimization of industrial thermal power systems. The objective functions of upper and lower levels are approximated by artificial neural network (ANN) models and the lower-level problem is analytically embedded through Karush-Kuhn-Tucker (KKT) optimality conditions. The reformulated single level optimization framework integrating ANN models and KKT constraints (ANN-KKT) is validated on benchmark problems and on real-world power generation operation of 660 MW coal power plant and 3
arXiv:2602.13746v1 Announce Type: new Abstract: Industrial thermal power systems have coupled performance variables with hierarchical order of importance, making their simultaneous optimization computationally challenging or infeasible. This barrier limits the integrated and computationally scaleable operation optimization of industrial thermal power systems. To address this issue for large-scale engineering systems, we present a fully machine learning-powered bi-level optimization framework for data-driven optimization of industrial thermal power systems. The objective functions of upper and lower levels are approximated by artificial neural network (ANN) models and the lower-level problem is analytically embedded through Karush-Kuhn-Tucker (KKT) optimality conditions. The reformulated single level optimization framework integrating ANN models and KKT constraints (ANN-KKT) is validated on benchmark problems and on real-world power generation operation of 660 MW coal power plant and 395 MW gas turbine system. The results reveal a comparable solutions obtained from the proposed ANN-KKT framework to the bi-level solutions of the benchmark problems. Marginal computational time requirement (0.22 to 0.88 s) to compute optimal solutions yields 583 MW (coal) and 402 MW (gas turbine) of power output at optimal turbine heat rate of 7337 kJ/kWh and 7542 kJ/kWh, respectively. In addition, the method expands to delineate a feasible and robust operating envelope that accounts for uncertainty in operating variables while maximizing thermal efficiency in various scenarios. These results demonstrate that ANN-KKT offers a scalable and computationally efficient route for hierarchical, data-driven optimization of industrial thermal power systems, achieving energy-efficient operations of large-scale engineering systems and contributing to industry 5.0.
Executive Summary
The article presents a novel data-driven bi-level optimization framework for industrial thermal power systems, leveraging artificial neural networks (ANNs) and Karush-Kuhn-Tucker (KKT) optimality conditions. This approach addresses the computational challenges associated with optimizing coupled performance variables of varying importance in large-scale systems. The proposed ANN-KKT framework is validated through benchmark problems and real-world applications in a 660 MW coal power plant and a 395 MW gas turbine system. The results demonstrate comparable solutions to traditional bi-level optimization methods, with significantly reduced computational time. The framework also delineates a feasible operating envelope that maximizes thermal efficiency while accounting for uncertainties, contributing to energy-efficient operations and Industry 5.0.
Key Points
- ▸ Introduction of a machine learning-powered bi-level optimization framework for thermal power systems.
- ▸ Use of ANNs to approximate objective functions and KKT conditions to embed lower-level problems.
- ▸ Validation through benchmark problems and real-world applications in coal and gas turbine systems.
- ▸ Achievement of optimal power output with minimal computational time.
- ▸ Delineation of a feasible operating envelope that maximizes thermal efficiency under uncertainties.
Merits
Innovative Approach
The integration of ANNs and KKT conditions provides a novel and scalable solution to the computational challenges of bi-level optimization in thermal power systems.
Real-World Applicability
The framework is validated on real-world systems, demonstrating its practical utility and effectiveness in industrial settings.
Efficiency and Speed
The ANN-KKT framework significantly reduces computational time, making it feasible for large-scale and real-time optimization tasks.
Demerits
Generalizability
The study's focus on specific power plants may limit the generalizability of the findings to other types of thermal power systems.
Data Dependency
The effectiveness of the ANN models is highly dependent on the quality and quantity of the training data, which may not always be readily available.
Complexity
The integration of ANNs and KKT conditions adds complexity to the optimization process, which may require specialized expertise for implementation and maintenance.
Expert Commentary
The article presents a significant advancement in the field of industrial thermal power system optimization. The integration of ANNs and KKT conditions offers a robust and scalable solution to the complexities of bi-level optimization. The validation on real-world systems underscores the practical relevance of the framework, demonstrating its potential to enhance energy efficiency and operational performance. However, the study's focus on specific power plants and the dependency on high-quality data highlight areas for further research. The framework's ability to delineate a feasible operating envelope under uncertainties is particularly noteworthy, as it addresses a critical need in industrial operations. Overall, the ANN-KKT framework represents a promising direction for achieving energy-efficient and sustainable industrial processes, aligning with the goals of Industry 5.0.
Recommendations
- ✓ Further validation of the ANN-KKT framework on a broader range of thermal power systems to assess its generalizability.
- ✓ Investigation into the robustness of the framework under varying data quality and quantity conditions to ensure its reliability in different operational scenarios.
