应用神经网络信息融合技术快速预测储层敏感性

快速、准确地诊断与预测储层敏感性损害问题一直是储层保护问题研究的一个重要领域。同时预测储层敏感性损害也是一门需要处理大量数据与信息的技术。信息融合技术是将各种途径、任意时间和任意空间上获得的信 息作为一个整体进行综合分析处理的技术。利用信息融合技术进行敏感性损害预测能够尽可能多地使用已有的数据获取最为准确的结果，在输入参数较少的情况下给出一个可靠的数值结果，且受人为因素干扰较少。基于 政进算法编制的神经网络信息融合技术储层敏感性快速预测软件分析表明，该方法受人为因素干扰小，可以渗断储层中邶种敏感性是主要因素并给出一个确切数值，同系统所需参数少，结果可靠（总体符合率达到91％），是一种能适用于现场的快速有效的方法，为油田合理处理敏感性损害提供了理论依据。

Study on Fast Diagnosis and Prediction of the Reservoir Damage by Using Neural Network Information Fusion Technique

How to fast and exact diagnose and predict reservoir damage problems are always one important area of reservoir protection study, and the prediction reservoir sensibility damage is also a technique, which needs deal with a block of data and information. Information fusion technique can analyze and deal with the information, which is collect by any methods and any time anywhere, from the whole. The method, which using the information fusion technique predict the sensibility, can make the most use of the input data to get the accuracy results, give the numerical results based on the less input data, further more, the method is not easily interrupted by man. The study show that based on the improved BP neural network information fusion method lessen the influence of man induced factors, need lesser parameter, it also can diagnose which sensibility is the main influent factor and the damage degree. The result is more credible (the total coincidence ratio is higher up 91%). This method is a real-time technique, and it is a suit to use in the oil field. It provides a theoretic support for the engineer deal with reservoir sensibility damage.