Modeling Method of Oilfield Mechanical Production Parameters Based on Unscented Particle Filter Neural Network

Abstract

The invention provides an oilfield mechanical oil extraction parameter modeling method based on an unscented particle filtering neural network. The method comprises the steps of determining efficiency affecting factors and performance variables in the oil extraction process of oilfield machinery; performing dimension reduction processing on load variables in a sample so as to reconstruct a new sample, and normalizing the new sample; building a neural network model based on the normalized new sample; estimating an optimal state variable of the neural network model by using an UPFNN algorithm, and building an oilfield mechanical oil extraction process model by using the optimal state variable; inputting X^ in the normalized new sample into the oilfield mechanical oil extraction process model to acquire a prediction result, comparing the prediction result with Y^ in the normalized new sample, wherein the oilfield mechanical oil extraction process model is effective if the comparison result is less than a preset error value, otherwise repeating all of the above steps until the comparison result is less than the preset error value. According to the invention, working conditions of the oilfield machinery are predicted through mining production laws of the oilfield machinery, and a basic model is provided for mining optimal production operations of the oilfield machinery.

Description

technical field [0001] The invention relates to the technical field of oilfield mechanical recovery, and more specifically, relates to a modeling method of oilfield mechanical recovery parameters based on an unscented particle filter neural network. Background technique [0002] Oilfield mechanical oil recovery is a mechanical oil recovery method, which is mainly composed of three parts: electric motor, ground transmission equipment and downhole pumping equipment. The oil production process of the oil field machine is mainly divided into two strokes, up and down. The up stroke, that is, the suspension point of the donkey head moves upward, and the sucker rod column and the liquid column need to be lifted, and the motor needs to consume a lot of energy; the down stroke, that is, the suspension point of the donkey head. Moving downwards, the oil field machine rod rotates and pulls to do work on the motor. During the up and down movement of the rod column, the load of the liqu...

AI Technical Summary

Problems solved by technology

During the up and down movement of the rod column, the load of the liquid column changes periodically, which makes the oilfield machine system consume a lot of energy in terms of motor work and transmission devices, resulting in low system efficiency and it is difficult to analyze the law of the oilfield machine process

Images