卡尔曼滤波与离散型数值积分相结合的示功图位移计算研究

Based on the Kalman Filter Dynamometer Displacement Is Calculated Research

为解决利用加速度传感器计算油井示功图的难题,提出一种卡尔曼滤波算法与离散型数值积分算法相结合用于油井示功图的位移计算和析出位移周期。卡尔曼滤波算法既过滤了加速度信号的噪声干扰,又保持了加速度信号的原始形状特征。位移周期的准确析出保证了位移计算原始数据的正确性。离散型数值积分算法可使位移测量的相对误差小于1%,满足了示功图精度的要求,适用于不同冲次的井型。

To solve the acceleration sensor wells dynamometer computing problem, we propose a Kalman filter algorithms and discrete numerical integration algorithms combine and analyze the displacement period approach , apply this oil displacement dyna-mometer calculations.Kalman filter algorithm both filters acceleration signal noise, while maintains the original shape of the accel-eration signal characteristics.Precipitation accurate displacement period to ensure the correctness of displacement calculation of the original data, using discrete numerical integration algorithm allows the relative displacement measurement error is less than 1%, to meet the requirements of precision dynamometer for pumping wells of different type .