Solver parameter influence on the results of multilayer perceptron for estimating power output of a combined cycle power plant

  • 1 Faculty of Engineering, University of Rijeka, Croatia


Previous work has determined the ability of using the Multilayer Perceptron (MLP) type of Artificial Neural Network (ANN) to estimate the power output of a Combined Cycle Power Plant (CCPP) in which optimization did not focus on the solver parameter optimization. In previous work, the solvers used the default parameters. Possibility exists that optimizing solver parameters will net better results. Two solver algorithm’s parameters are optimized: Stochastic Gradient Descent (SGD) and Adam, with 140 and 720 parameter combinations respectively. Solutions are estimated through the use of Root Mean Square Error (RMSE). Lowest RMSE achieved is 4.275 [MW] for SGD and 4.259 [MW] for Adam, achieved with parameters: = 0.05, = 0.02, and nesterov=True for SGD and with parameters = 0.001, 1 = 0.95, 2 = 0.99, and amsgrad=False for Adam. Only a slight improvement is shown in comparison to previous results (RMSE=4.305 [MW]) which points towards the fact that solver parameter optimization with the goal of improving results does not justify the extra time taken for training.



  1. Mrzljak, V., Poljak, I., Orović, J., & Prpić-Oršić, J. (2019, January). Numerical analysis of real open cycle gas turbine. In INTERNATIONAL SCIENTIFIC CONFERENCE HIGH TECHNOLOGIES. BUSINESS. SOCIETY 2019.
  2. Mrzljak, V., AnĎelić, N., Lorencin, I., & Car, Z. (2019). Analysis of Gas Turbine Operation before and after Major Maintenance. Pomorski zbornik, 57(1), 57-70.
  3. Glazar, V., Mrzljak, V., & Gubic, T., (2019) Thermodynamic Analysis of Combined Cycle Power Plant. 14th International Conference Heat Transfer, Fluid Mechanics and Thermodynamics, 355-341
  4. Lorencin, I., AnĎelić, N., Mrzljak, V., & Car, Z. (2019). Genetic Algorithm Approach to Design of Multi-Layer Perceptron for Combined Cycle Power Plant Electrical Power Output Estimation. Energies, 12(22), 4352.
  5. Lorencin I., Car Z., Kudlaček J., Mrzljak V., AnĎelić N, & Blažević S. (2019). Estimation of Combined Cycle Power Plant Power Output Using Multilayer Perceptron Variations, 10th International Technical Conference - Technological Forum 2019 – Proceedings, 94-98
  6. Lorencin, I., AnĎelić, N., Mrzljak, V., & Car, Z. (2019). Marine Objects Recognition Using Convolutional Neural Networks. NAŠE MORE: znanstveno-stručni časopis za more i pomorstvo, 66(3), 112-119.
  7. Baressi Šegota, S., AnĎelić, N., Kudláček, J., & Čep, R. (2019). Artificial neural network for predicting values of residuary resistance per unit weight of displacement. Pomorski zbornik, 57(1), 9-22.
  8. AnĎelić, N., Blažević, S., & Car, Z. (2018, January). Trajectroy Planning Using Genetic Algorithm For Three Joints Robot Manipulator. In International Conference on Innovative Technologies, IN-TECH 2018.
  9. Blažević, S., AnĎelić, N., & Car, Z. (2017, January). Research of Unstable Behavior of Iterative Path Planning Algorithm for Robot Manipulator. In IN-TECH 2017 International Conference on Innovative Technologies.
  10. Lorencin, I., AnĎelić, N., Španjol, J., & Car, Z. (2020). Using multi-layer perceptron with Laplacian edge detector for bladder cancer diagnosis. Artificial Intelligence in Medicine, 102, 101746.
  11. Baressi Šegota, S., Lorencin, I., Ohkura, K., & Car, Z. (2019). On the Traveling Salesman Problem in Nautical Environments: an Evolutionary Computing Approach to Optimization of Tourist Route Paths in Medulin, Croatia. Pomorski zbornik, 57(1), 71-87.
  12. Mrzljak, V., Blecich, P., AnĎelić, N., & Lorencin, I. (2019). Energy and Exergy Analyses of Forced Draft Fan for Marine Steam Propulsion System during Load Change. Journal of Marine Science and Engineering, 7(11), 381.
  13. Lorencin, I., AnĎelić, N., Mrzljak, V., & Car, Z. (2019). Exergy analysis of marine steam turbine labyrinth (gland) seals. Pomorstvo, 33(1), 76-83.
  14. Mrzljak, V., Car, Z., Kudláček, J., AnĎelić, N., Lorencin, I., & Blažević, S. (2019) Analysis of Two Methods For Steam Turbine Developed Power Calculation In Industry 4.0. 10th International Technical Conference - Technological Forum 2019 – Proceedings, 103-110
  15. Friedman, J., Hastie, T., & Tibshirani, R. (2001). The elements of statistical learning (Vol. 1, No. 10). New York: Springer series in statistics.
  16. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT press.
  17. Bishop, C. M. (2006). Pattern recognition and machine learning. springer.
  18. Bottou, L. (2010). Large-scale machine learning with stochastic gradient descent. In Proceedings of COMPSTAT'2010 (pp. 177-186). Physica-Verlag HD.
  19. Dozat, Timothy. Incorporating nesterov momentum into adam. (2016). In Proceedings of 2016 International Conference of Learning Representations (pp. 1-4)
  20. Kingma, D. P., & Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
  21. Reddi, S. J., Kale, S., & Kumar, S. (2019). On the convergence of adam and beyond. arXiv preprint arXiv:1904.09237.
  22. Géron, A. (2019). Hands-On Machine Learning with Scikit- Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems. O'Reilly Media.
  23. Tüfekci, P. (2014). Prediction of full load electrical power output of a base load operated combined cycle power plant using machine learning methods. International Journal of Electrical Power & Energy Systems, 60, 126-140.

Article full text

Download PDF