Pipeline leakage incidents have led to serious environmental concerns, prompting the development of various methods to detect leakage in pipelines. Hence, there is a justified need for early leakage detection as well as leakage prediction system. Many of the leak detection techniques are based on computational intelligence techniques such as artificial neural networks, support vector machine, and fuzzy systems. This paper focuses on determining leak detection parameters such as mass flow rate, temperature, and pressure with and without a leak in pipelines evaluated using OLGA multiphase software. An OLGA-based computerized model is used in leak simulation for analyzing inlet and outlet parameters such as mass flow rate, temperature, and pressure over the flow inside the pipeline. The leak sizes were varied from 0% to 50 % leak opening and the inlet and outlet parameters were measured and studied. The pressure and mass flow rates are observed to decrease with increasing leak size, while temperature decreases with leak size until 25 % and later increases. Mass flow rate is observed to be the most important parameter in detecting a leak and localizing it. The maximum percentage of variation in mass flow rate was observed to be 33.6 % for 50 % leak openings, for a single leak, and 32.4 % for multi-leak scenario.

管道泄漏事故引发了严重的环境问题，促使开发各种检测管道泄漏的方法。因此，对早期泄漏检测以及泄漏预测系统的需求是合理的。许多泄漏检测技术都基于计算智能技术，例如人工神经网络、支持向量机和模糊系统。本文侧重于确定泄漏检测参数，例如使用 OLGA 多相软件评估的管道中有无泄漏的质量流量、温度和压力。在泄漏模拟中使用基于 OLGA 的计算机模型来分析入口和出口参数，例如管道内流动的质量流量、温度和压力。泄漏尺寸从 0% 到 50% 的泄漏开口不等，并且测量和研究了入口和出口参数。观察到压力和质量流量随着泄漏尺寸的增加而降低，而温度随着泄漏尺寸的增加而降低，直到 25% 和以后增加。质量流量被认为是检测泄漏和定位泄漏的最重要参数。对于 50% 的泄漏开口、单次泄漏和多泄漏情况，质量流量变化的最大百分比观察到为 33.6%。