Comparison of Machine Learning Algorithms for Structure State Prediction in Operational Load Monitoring

Abstract

The aim of operational load monitoring is to make predictions about the remaining usability time of structures, which is extremely useful in aerospace industry where in-service life of aircraft structural components can be maximized, taking into account safety. In order to make such predictions, strain sensors are mounted to the structure, from which data are acquired during operational time. This allows to determine how many load cycles has the structure withstood so far. Continuous monitoring of the strain distribution of the whole structure can be complicated due to vicissitude nature of the loads. Sensors should be mounted in places where stress and strain accumulations occur, and due to experiencing variable loads, the number of required sensors may be high. In this work, different machine learning and artificial intelligence algorithms are implemented to predict the current safety factor of the structure in its most stressed point, based on relatively low number of strain measurements. Adaptive neuro-fuzzy inference systems (ANFIS), support-vector machines (SVM) and Gaussian processes for machine learning (GPML) are trained with simulation data, and their effectiveness is measured using data obtained from experiments. The proposed methods are compared to the earlier work where artificial neural networks (ANN) were proven to be efficiently used for reduction of the number of sensors in operational load monitoring processes. A numerical comparison of accuracy and computational time (taking into account possible real-time applications) between all considered methods is provided.

Keywords: ANFIS; gaussian process regression; hat-stiffened panel; operational load monitoring; strain measurement; structural health monitoring; support-vector machine.

Figure 1

Considered aerostructure: (a) photograph, (b) top strain gauges SG1–SG3 positions and load points LC1–LC3, (c) bottom strain gauges SG4–SG6 positions, (d) finite element model and boundary conditions. Reprinted from [15].

Figure 2

Experimental testing. Reprinted from [15].

Figure 3

Generated data samples of ISF (inverse safety factor) for example force value of 40 N.

Figure 4

Inputs and outputs of the prediction model.

Figure 5

Algorithm for finding best hyperparameters values for each model.

Figure 6

Algorithm for finding best suited models of each type.

Figure 7

Structure of ANFIS.

Figure 8

Results of testing ANFIS models: (a) load case LC1, numerical input data; (b) load case LC1, experimental input data; (c) load case LC2, numerical input data; (d) load case LC2, experimental input data; (e) load case LC3, numerical input data; (f) load case LC3, experimental input data.

Figure 9

Results of testing SVM models: (a) load case LC1, numerical input data; (b) load case LC1, experimental input data; (c) load case LC2, numerical input data; (d) load case LC2, experimental input data; (e) load case LC3, numerical input data; (f) load case LC3, experimental input data.

Figure 10

Graphical comparison of accuracy and computational time for obtained SVM models.

Figure 11

Results of testing GPR models: (a) load case LC1, numerical input data; (b) load case LC1, experimental input data; (c) load case LC2, numerical input data; (d) load case LC2, experimental input data; (e) load case LC3, numerical input data; (f) load case LC3, experimental input data.

Figure 12

Graphical comparison of accuracy and computational time for obtained GPR models.

Figure 13

Results of testing chosen models of different methods: (a) load case LC1, numerical input data; (b) load case LC1, experimental input data; (c) load case LC2, numerical input data; (d) load case LC2, experimental input data; (e) load case LC3, numerical input data; (f) load case LC3, experimental input data.

Figure 14

Graphical comparison of accuracy and computational time for chosen models.

Figure 15

Results of testing chosen models with experimental input from Supplementary S8 of [15]: (a) time plot of the applied force; (b) time plots of the ISF obtain from different models.

Figure 16

Histograms of RMSE for training 1000 models: (a) ANF1, SVM2, and GPR2 models, (b) ANN models.

Figure 17

Results of t-test repeatability for chosen models.