سامانه همایش‌ها TSTA | ثبت‌نام و ارسال مقاله کنفرانس‌ها و کنگره‌های ملی

صفحه اصلی / دومین کنفرانس ملی عصر انفجار تکنولوژی؛ هوش مصنوعی، تحولی در صنعت، تجارت و زنجیره تامین و دومین کنفرانس ملی علم داده در کاربردهای مهندسی

Designing a Machine Learning Model with LSTM and CNNs to Make the Quality Control Process of Liquefied Gas Tankers Intelligent

نویسندگان :

Raha Pakzad¹

1- هلدینگ صنعتی مارال

کلمات کلیدی :

Liquefied gas،Machine learning،LPG،quality control (QC)،Inteligen model

چکیده :

Transporting liquefied gas (like LNG or LPG) by tanker is a complex and hazardous process. Because of the flammable and pressurized nature of the gas, any fault can lead to serious safety risks, environmental damage, or legal violations. That’s why strict quality control (QC) is essential to ensure everything operates safely, efficiently, and within regulatory guidelines.Traditionally, QC has been done through manual inspections and fixed threshold-based systems. For example, if a pressure or temperature reading goes above or below a set limit, an alarm might go off. However, these methods have limitations—they depend on human judgment, are often slow, and can miss subtle warning signs. Static thresholds also don’t adapt to changing conditions or recognize patterns over time.To solve these issues, the study proposes a machine learning (ML) model that automates and improves the QC process. Instead of relying only on fixed rules, the model learns from real-world data—like sensor readings (e.g., pressure, temperature), operational conditions (e.g., load levels, travel routes), and historical maintenance records (e.g., past faults, repairs). By analyzing all this data, the ML model can detect patterns and predict potential failures or quality problems before they happen. This allows tanker operators to take preventive action, improving safety and avoiding costly downtime. When tested, the model proved to be accurate and reliable across different scenarios. It reduced unplanned maintenance, improved safety margins, and showed strong potential for making liquefied gas transport more intelligent, efficient, and secure