Optimization method of decision-making parameters in oilfield mechanical recovery process based on preference multi-objective optimization

The invention provides an oil field mechanical extraction process decision parameter optimization method based on preference multi-objective optimization. The oil field mechanical extraction process decision parameter optimization method based on preference multi-objective optimization includes the steps: determining the efficiency influence factor and the performance variable during the oil extraction process of an oil field machine; performing dimension reduction processing on the load variable in a sample, constructing a new sample, and normalizing the new sample; based on the normalized new sample, constructing a neural network model; utilizing an ST-UPFNN algorithm to estimate the optimal state of the state variable formed by the weight thresholds in the neural network model; utilizing the optimal state variable to reconstruct the updated neural network model to obtain an oil extraction process model for an oil field machine; constructing a preference function for the practical liquid production capacity; utilizing a multi-objective evolutionary algorithm to optimize the top and bottom limitation of each decision parameter; and plugging the optimized decision variable into the oil extraction process model for an oil field machine, calculating the average value of the system performance of the optimized decision variable, and comparing with the average value of the system performance of the practical sample. The oil field mechanical extraction process decision parameter optimization method based on preference multi-objective optimization can improve the production efficiency for oil extraction of an oil extraction machine, and can reduce energy consumption.