Energy modulators (EM), also known as energy absorbers, are safety-critical components used to control shocks and movements in the load path. EMs are textile devices typically manufactured from nylon, Kevlar® and other materials, and control loads by breaking the rows of stitches that bind together a strong base webbing as shown in Figure 1. has gone To prevent injury from shock loads when the harness prevents a fall. EMs are also widely used in parachute systems to control shock loads during various stages of parachute system deployment.
Random Forest is an advanced algorithm for data classification used in statistics and machine learning. It is an easy-to-use and highly flexible learning method. The Random Forest algorithm is capable of modeling both discrete and continuous data and can handle large data sets, making it applicable in many situations. It also makes it easier to estimate the relative importance of variables and maintains accuracy even when values are missing in the data set.
Random forests model the relationship between a response variable and a set of predictor or independent variables by building a collection of decision trees. Each decision tree is constructed from a random sample of data. The individual trees are then combined using methods such as averaging or voting to determine the final prediction (Figure 2). A decision tree is a non-parametric supervised learning algorithm that partitions data using a series of branching binary decisions. Decision trees inherently identify key features of data and provide a ranking of the contribution of each feature based on when it becomes relevant. This capability can be used to determine the relative importance of input variables (Figure 3). Decision trees are useful for exploring relationships but their accuracy may deteriorate unless they are combined with random forests or other tree-based models.
The performance of a random forest can be evaluated using out-of-bag error and cross-validation techniques. Random forests often use random sampling with replacement from the original dataset to build each decision tree. This is also known as bootstrap sampling and creates a bootstrap forest. Data included in the bootstrap sample is called in-the-bag, while data not selected is out-of-bag. Since the out-of-bag data were not used to build the decision tree, they can be used as an internal measure of model accuracy. Cross-validation can be used to evaluate how well the results of a random forest model will fit an independent dataset. In this approach, the data is divided into a training dataset used to develop the decision trees and build the model and a validation dataset used to evaluate the performance of the model. Evaluation of the model on an independent validation dataset provides an estimate of how well the model will perform in practice and helps avoid problems such as overfitting or sampling bias. A good model performs well.
Both training data and validation data.
The complex nature of the EM system made it difficult for the team to identify how different parameters affected EM behavior. A bootstrap forest analysis was applied to the test dataset and was able to identify five key variables associated with high probability of loss and/or abnormal behavior. The identified key variables provided the basis for further evaluation and redesign of the EM system. These findings also provided essential insights for investigations and aided in the development of flight reasoning for future use cases.
Contact Dr. Sarah R. Wilson for information. sara.r.wilson@nasa.gov