Enhanced PSO-Integrated Machine Learning for Launch Vehicle Engine Temperature and Pressure Monitoring for Risk Prediction

Maintaining the health of the launch vehicle engines is essential to the mission’s success, including safe ascent, payload deployment, and orbital insertion. Preemptive maintenance is made possible by early anomaly identification, which reduces mission risks and maximizes operational effectiveness. Engineers can prevent catastrophic failures by proactively mitigating possible difficulties by accurate forecasting before launch. During space exploration missions, timely maintenance and modifications protect personnel, spacecraft, and important payloads, further enhancing mission reliability. This Research work presents a pioneering approach to predict launch vehicle engine health, the study integrates Particle Swarm Optimization (PSO) with Enhanced PSO (EPSO) to precisely determine vehicle engine health. Employing real vibration data from rocket launches, and developing various machine learning models such as Random Forest, Support Vector Machine, Naïve Bayes, and k-Nearest Neighbours(KNN) among the four models we chose KNN since it showed better accuracy. We integrated the EPSO to the KNN model to implement the accuracy for better prediction. This research contributes significantly to advancing health diagnostics and optimization for launch vehicles, ultimately bolstering the safety and dependability of aerospace systems.