基于逻辑回归的电缆吸附卡风险评价方法

针对电缆测井现场无法预测电缆吸附卡的情况，基于逻辑回归建立一种电缆吸附卡风险评价方法，再使用Monte-Carlo进行模拟，得到电缆吸附卡风险概率。首先，根据现场施工经验，建立电缆吸附卡事故树，明确影响电缆吸附卡的原因，确定电缆吸附卡的主控因素；然后，构建基于逻辑回归的电缆吸附卡风险评价模型，带入训练集中井的主控因素对模型进行训练，提高模型的可靠度；最后，采用Monte-Carlo方法进行模拟，计算电缆吸附卡的风险概率，实现对电缆吸附卡的风险判断，以新疆某油田32口井为训练集对同区域的两口井进行了实例分析，评价结果与现场相符。结果表明：所建立的方法能够较好地利用大数据对电缆吸附卡风险进行定量评价，对于保障电缆测井作业安全和实现智能化具有一定理论指导意义，是一种新颖科学的评价方法。

Risk assessment cable sticking based on Logistic regression

Aiming at the situation that the cable sticking cannot be predicted in the field of wireline logging, an evaluation method of cable sticking is established based on logistic regression, and then Monte Carlo simulation is used to obtain the probability of cable sticking. Firstly, according to the on-site construction experience, the cable sticking accident tree was established to clarify the reasons affecting the cable sticking and determine the main control factors of the cable sticking. Then, the cable sticking evaluation model based on Logistic regression was constructed, and the main control factors of the training well were used to train the model to improve the reliability of the model. Finally, Monte Carlo method is used to calculate the risk probability of cable sticking, and realize the risk judgment of cable sticking cable. Taking 32 Wells in an oilfield in Xinjiang as the training set, two Wells in the same area are analyzed, and the evaluation results are consistent with the field. The results show that the method established in this paper can make good use of big data to quantitatively evaluate the risk of cable sticking, and has certain theoretical significance for ensuring the safety of wireline logging operation and realizing intelligence. It is a novel and scientific evaluation method.