Loading...

Implementation of the Lubricant Analysis Program Using Artificial Intelligence to Evaluate the Condition of Dump Truck Diesel Engine

Massoudi, Mohammad Amin | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: English
  3. Document No: 57731 (58)
  4. University: Sharif University of Technology, International Campus, Kish Island
  5. Department: Science and Engineering
  6. Advisor(s): Behzad, Mehdi; Moradi, Reza
  7. Abstract:
  8. Dump trucks are one of the most important and valuable equipment used in mining, and the implementation of proper maintenance and monitoring programs for these machines is of special significance. The most critical part of a dump truck is its diesel engine, and the most effective method for monitoring its condition is to conduct lubricant analysis. Accuracy and speed in the lubricant analysis process significantly contribute to enhancing the efficiency of maintenance and monitoring programs for the equipment. Given the complexities of analyzing the condition of diesel engines in dump trucks due to the interactions among multiple engine parameters and various lubricant characteristics, there is a possibility of incorrect condition assessment by experts. Nowadays, the development of artificial intelligence methods and their application in various engineering problems has garnered significant attention from researchers due to their high capability in solving complex and nonlinear issues. In this regard, the goal of this project is to implement a lubricant analysis program using artificial intelligence to assess the condition of diesel engines in dump trucks. To achieve this goal, 11,264 lubricant analysis samples from dump truck diesel engines were collected. Subsequently, two artificial intelligence models, Feedforward Neural Network (FNN) and Convolutional Neural Network (CNN), were selected based on the type of data, and suitable features were chosen to serve as inputs for the models, to learn the analysis. Subsequently, by introducing innovations and applying them to the AI models to enhance accuracy and quality, advanced models were developed that significantly outperformed the initial model in lubricant analysis. The first innovation involved implementing a specific normalization tailored to the wear regime of each machine model based on the results of key tests, which was introduced as a new input. The second innovation leveraged the impact of the relationships between the states of the analytical sections during the training and testing processes. The developed model achieved an accuracy of approximately 97% during the training phase and 93% in predicting test samples. After developing and evaluating the FNN and CNN models for intelligent analysis of industrial lubricant test results, software with a user interface was designed and implemented. The developed software was subsequently validated with industrial data, and its results indicate that the predictions of the developed models were similar to, and in some cases more accurate than, expert assessments. Therefore, the use of the developed models can practically increase the quality, accuracy, and speed of the lubricant analysis process, leading to more precise and timely decision-making in the industry
  9. Keywords:
  10. Condition Monitoring ; Oil Analysis ; Intelligence Operator ; Neural Network ; Diesel Engines ; Dump Truck Diesel Engine

 Digital Object List

 Bookmark

No TOC