Prediction of the Remaining Recoverable Reserves of Oil Well before Refracturing Based on Grey Model Improved by Particle Swarm Optimization

In order to obtain the remaining recoverable reserves of refract ring well, based on grey system theory, we applied the forward and backward difference to the white differential equation of the grey model and introduced a parameter. Then an improved grey model was established, accordingly, the original expression of the background value with the introduced parameter was derived. On this basis, the particle swarm optimization algorithm was applied to search the best parameter. Ultimately, we used the improved grey model and combined with the daily oil production of the oil well before refract ring to forecast the remaining recoverable reserves. The prediction result indicated that the modified grey model was much more accurate than the unimproved grey model. In addition, it was simple to calculate and demands just a few raw data. Therefore, it could be used as another prediction method apart from material balance method.