Selecting the most effective welding settings impacts mechanical strength and weld quality, with parameters like current, voltage, and speed playing pivotal roles. The mechanical part encompasses material properties, welding process variables, and structural integrity, all contributing to the overall weld quality and strength. By integrating these mechanical factors with predictive modelling, a comprehensive understanding of weld performance can be achieved, enabling optimized welding settings and enhanced weld quality assurance. This study assesses and compares machine learning algorithms such as a random tree, random forest, and C4.5 to determine their predictive capability regarding the tensile strength in Monel 400 Weldments. By utilizing a dataset comprising 32 instances with attributes like Current, Voltage, and Speed, models were developed and assessed using K-Fold cross-validation. Among these algorithms, the random tree models emerge as the most proficient in accurately predicting the tensile strength for Monel 400 Weldments through classification ML techniques. Similarly, regression algorithms have been deployed to assess the dataset by varying the train-test split ratio and gradient boosting, which exhibited superior performance with a higher R2 value of 0.99. Both random tree and Gradient boosting algorithms have commonly been recommended, with current being the most influential factor affecting tensile strength.
1 December 2024 in Recent Advances in Electrical & Electronic Engineering (Formerly Recent Patents on Electrical & Electronic Engineering)
Background: Consumption of electricity always varies based on demand. The load cluster pattern aims at categorizing periodical changes over a specific time. Predicting the electric load was the initial goal of this study. Additionally, the outcomes of the load prediction were utilized as data for categorizing electrical loads using a descriptive-analytical method. Objective: The study has dealt with a matching of load-side electric demand with the electric supply. To ensure dependable power-generating stability, it is vital to anticipate and categorize loads. Thus, the research presented here has focused on electrical load forecasting and classification. Methods: Alternative algorithms, including Naive Bayes, decision tree, and support vector machine classifier, were employed to address the cluster pattern. The data used for this research presentation was collected from the D Block of the Kamaraj College of Engineering and Technology, K. Vellakulam, India, every 15 minutes. Multiple unsuitable loaded circumstances were ignored during the pre-processing of the dataset. Additionally, other algorithms, like Naive Bayes, decision tree, and support vector machine, were used to categorize the raw data. The processing of data was done by a feature selection approach. Results: The performance was predicted by comparing the entire machine learning algorithms. Out of the machine learning techniques, an accuracy of 4.2% for Academic Block 4, a precision of 33% for Boys Hostel, a recall score of 4.7% for Academic Block 4, and an F1 score of 5.3% for Academic Block 4, were obtained. Conclusion: In the study, the decision tree algorithm has shown promising performance than the other machine learning techniques used.
23 December 2024 in Environment Conservation Journal
In today's rapidly evolving technological world, collecting and analyzing the environmental data has become crucial for ensuring safety and informed decision-making. By analyzing the collected data, people can better plan, manage risks, and respond proactively to changing conditions. Our work aims to provide a comprehensive system for monitoring and analyzing diverse ecological conditions to support land investment and construction decisions. Our innovative system closely observes the environmental conditions in a given area to ensure overall safety and well-being. By deploying ecosentinel, which is an array of sensors, diverse ecological conditions encompassing sound, temperature and gas levels are meticulously monitored. This comprehensive data is then securely stored within the cloud infrastructure, where it undergoes rigorous analysis utilizing historical records for visualization. Through continuous sensor data uploads to the cloud, a well-structured analysis is periodically conducted, presenting the information in a visually accessible manner. To gain valuable insights from the data recorded over a month, a statistical and correlation analysis is conducted. This analysis is essential to understand the relationships and patterns within the recorded values, which can help in making informed decisions and identifying any significant trends or anomalies in that specific area. Sound, temperature, and gas sensors have maximum coverage of 20 m, 10 m, and 50 m respectively, based on sensitivity, environment, and placement. The system offers real-time snapshots and predictive insights into environmental conditions. It highlights significant trends and correlations, aiding in proactive decision-making and risk assessment for investments and construction projects. Ecosentinel empowers users to evaluate pollution levels and forecast environmental scenarios without extensive research or site visits. It is a valuable tool for individuals and organizations aiming for informed and sustainable planning in the region. Ensuring the accuracy and reliability of the system’s alerts and insights, particularly in the face of dynamic and unpredictable environmental conditions remains a complex task.