广义预测控制在煤层气水平井中的应用

Study on application of generalized predictive control in coal seam gas horizontal well

针对目前煤层气水平井开发系统存在滞后、惯性以及时变等突出问题,提出了一种基于卡尔曼(CARIMA)模型的广义预测控制方法。根据煤层气水平井开采中煤层气的储存特征,在井下配置多个流量控制阀(ICV)以及实时物理测量传感器,采集如压力、温度、流量等煤层气水平井动态特征参数。将广义预测控制(GPC)应用于ICV自动调节水平井渗流系统内指定的压力。仿真结果表明:GPC算法控制有效,当噪声或异常值出现时,验证了算法的鲁棒性。

Since there are serious time delay, inertia and time varying existing in the development system of the coal seam gas horizontal well, the paper proposed a generalized predictive control (GPC) method based on CARIMA model. According to the storage characteristics of the coal seam gas during development of the coal seam gas horizontal well, multiple inflow control valves (ICV) and real-time physical measurement sensors were equipped in the underground to collect the dynamic characteristic parameters of the coal seam gas horizontal well such as pressure, temperature and flow. The GPC was applied to ICV to automatically regulate the pressure at specified value in the flow system. Simulation results showed: the GPC method worked effectively, the robustness of the method was verified while noise or outliers occurred.