Research and Application of Multi-Fusion Algorithm in Equipment Safety Hazard Prediction and Analysis Scenarios

In the current era, risk assessment and hazard identification (HAZID) are significant for the safe and efficient production of the coal chemical industry (methanol to olefins). Nevertheless, none of the conventional studies focused on HAZID assessment, which is crucial to diminishing the risk of accidents. Therefore, this paper proposes an effective 2LPIME and EDRc-FLS-based defect root cause analysis (RCA) and hazard assessment in equipment safety. Initially, sensor data are collected, followed by pre-processing. Next, the features are extracted and fed into the proposed Zakharov Exponential Fractional SiLU Long Short-Term Memory (ZEFSiLU-LSTM), where the defect is classified. Similarly, the sensor data are inputted to the thermal runaway detection. The thermal runaway is mitigated using the Beale function-PID Controller (BF-PID). Then, the processes like signal filtering, windowing, frequency-domain conversion and peak value analysis are done. The vibration is analyzed and handled via the Kalman filter based on the RMS value. Also, the thermal runaway mitigated data and vibration-handled data are fused and then the correlation is estimated. Based on the defect features and correlated data, the root cause is recognized. Further, the HAZID assessment is done to identify the low, medium and high risks. Overall, the proposed methodology had high robustness with 99.12% accuracy.