Optimization of a Jet Engine Compressor Disc with Application of Artificial Neural Networks for Calculations Related to Time and Mass Criteria

Abstract

The paper presents the results of a series of numerical research on the possibility of applying Artificial Neural Networks (ANNs) for the ultimate strength calculations of selected parts of rotating machines. The layout and the principle of the algorithm operation were described, beginning from the general assumptions and then moving on to the detailed description of the subsequent modules. The effects of applying the algorithm were presented on the example of the compressor disc analysis. The significant benefits of its application were the reduction of optimization time by about 40% and the disc weight reduction by 0.5 kg. The accuracy of ANNs was different in each iteration of the presented algorithm. Finally, high accuracy of neural networks was achieved with the following mean values of relevant indices reached in the last iteration: RMSE = 0.5983, MAPA = 0.0733 and R 2 = 0.99895. The further perspectives of the undertaken research were defined at the end.

Artificial Neural Network Structure Optimisation in the Pareto Approach on the Example of Stress Prediction in the Disk-Drum Structure of an Axial Compressor

“…It indicated that ANN is a tool capable of supporting or replacing classical numerical calculations. As a result of the preliminary numerical analysis presented in [10], the authors obtained knowledge in the field of optimal network selection. It provided knowledge on the network structure used.…”

“…In the analysed area of the literature, the authors did not come across any publications related to the use of an ANN in disk systems or disk-drum turbine engines optimisation. Previous attempts of the authors to use an ANN in compressor disc calculations are presented in publications [9,10]. This paper is an extension of these works, aimed at optimising and extending the research of the previously used algorithm.…”

“…ANNs are increasingly being used in aviation [8][9][10]. They are used in reliability analysis [11,12], fault detection [13][14][15][16], identification [17], control [18,19], and design and optimisation [20][21][22][23][24][25].…”

Method for Parameter Tuning of Hybrid Optimization Algorithms for Problems with High Computational Costs of Objective Function Evaluations

“…Nevertheless, part of the research is devoted not necessarily to the metamodels themselves, but to techniques that minimize the number of costly training samples needed to fit an efficient surrogate, or even to guide the sampling procedure based on areas of metamodel high uncertainty [16]-this technique is known as the infill process. Nonetheless, numerous papers utilize surrogates to replace costly computer simulations, e.g., as the authors of [17] minimized the mass of the jet engine components using ANN, which was carried out 40% faster than the optimization carried out on the numerical-analytical model. Moreover, the authors incorporated the infill process during the optimization.…”

Modification of Genetic Algorithm Based on Extinction Events and Migration

“…A mutation consists of negating a certain number of bits in selected individuals. Genetic algorithms tend to be connected with other optimization methods, such as a Tabu search [22], or light algorithms, such as fuzzy logic [23] or a neural network [24,25]. There have been few attempts to modify the genetic algorithm bases.…”