A Machine Learning Approach for Solid Rocket Motor Data Analysis and Virtual Sensor Development

Williams, Aimee; Himschoot, Austin; Saafir, Malik; Gatlin, Maia; Pendleton, David E.; Alvord, David

doi:10.2514/6.2020-3935

Cited by 5 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is an important contribution to the development of the rocket engine field. In 2020, Aimee Williams et al combined input measurement data and machine learning to create a "virtual sensor" that can provide critical information that traditional measurement methods cannot obtain due to the inability to place sensors in the combustion chamber [18]. In 2021, Dongxu Liu et al proposed and investigated a deep convolutional neural network (CNN) architecture that uses the finite element method (FEM) to generate labeled training data to evaluate the scale of defects in solid rocket engines coexisting with bore cracks and propellant debonding [19].…”

Section: Artificial Intelligence and Solid Rocket Enginesmentioning

confidence: 99%

A Chinese BERT-Based Dual-Channel Named Entity Recognition Method for Solid Rocket Engines

2023

View full text Add to dashboard Cite

With the Chinese data for solid rocket engines, traditional named entity recognition cannot be used to learn both character features and contextual sequence-related information from the input text, and there is a lack of research on the advantages of dual-channel networks. To address this problem, this paper proposes a BERT-based dual-channel named entity recognition model for solid rocket engines. This model uses a BERT pre-trained language model to encode individual characters, obtaining a vector representation corresponding to each character. The dual-channel network consists of a CNN and BiLSTM, using the convolutional layer for feature extraction and the BiLSTM layer to extract sequential and sequence-related information from the text. The experimental results showed that the model proposed in this paper achieved good results in the named entity recognition task using the solid rocket engine dataset. The accuracy, recall and F1-score were 85.40%, 87.70% and 86.53%, respectively, which were all higher than the results of the comparison models.

show abstract

Section: Artificial Intelligence and Solid Rocket Enginesmentioning

confidence: 99%

A Chinese BERT-Based Dual-Channel Named Entity Recognition Method for Solid Rocket Engines

2023

View full text Add to dashboard Cite

show abstract

“…Williams et al [21] proposed the concept of "virtual sensors", which combines machine learning with obtained measurement data to provide critical information that the sensor cannot be placed at key locations of rocket motors due to environmental factors. Lv et al [22] identified the state of liquid rocket motors by constructing a fusion recurrent convolutional neural network (FRCNN).…”

Section: Introductionmentioning

confidence: 99%

Lightweight RepVGG-Based Cross-Modality Data Prediction Method for Solid Rocket Motors

Yang,

Zheng,

Wang

et al. 2023

Sensors

View full text Add to dashboard Cite

Solid rocket motors (SRMs) have been popularly used in the current aerospace industry. Performance indicators, such as pressure and thrust, are of great importance for rocket monitoring and design. However, the measurement of such signals requires high economic and time costs. In many practical situations, the thrust measurement error is large and requires manual correction. In order to address this challenging problem, a lightweight RepVGG-based cross-modality data prediction method is proposed for SRMs. An end-to-end data prediction framework is established by transforming data across different modalities. A novel RepVGG deep neural network architecture is built, which is able to automatically learn features from raw data and predict new time-series data of different modalities. The effectiveness of the proposed method is extensively validated with the field SRM data. The accurate prediction of the thrust data can be achieved by exploring the pressure data. After calculation, the percentage error between the predicted data and the actual data is less than 5%. The proposed method offers a promising tool for cross-modality data prediction in real aerospace industries for SRMs.

show abstract

“…In our analysis, the data of solid rocket motor which can be considered as time series. Some common algorithms like CNN [12; 13], RNN [14], LSTM [15; 16], GRU [15] and transformers [17] have been applied for time series problems, which have great potential and promising prospect for industrial applications.…”

Section: Introductionmentioning

confidence: 99%

Deep learning-based total impulse prediction method in ignition process of solid rocket motor

Yang,

Wang,

Zheng

et al. 2023

International Workshop on Signal Processing and Machine Learning (WSPML 2023)

View full text Add to dashboard Cite

A Machine Learning Approach for Solid Rocket Motor Data Analysis and Virtual Sensor Development

Cited by 5 publications

References 4 publications

A Chinese BERT-Based Dual-Channel Named Entity Recognition Method for Solid Rocket Engines

A Chinese BERT-Based Dual-Channel Named Entity Recognition Method for Solid Rocket Engines

Lightweight RepVGG-Based Cross-Modality Data Prediction Method for Solid Rocket Motors

Deep learning-based total impulse prediction method in ignition process of solid rocket motor

Contact Info

Product

Resources

About