| High energy consumption and low overall system efficiency are the main issues appeared on the offshore petroleum platform. For addressing these problems, based on the offshore platform injection-production dynamic simulator, an intelligent model-free gradient efficiency optimization control method and a GA efficiency optimization control method based on the BP neural network model are designed in this paper, which realize the platform single-well and multi-well optimization control. Further offshore platforms integrated optimization control method are designed. Ultimately these optimization control method are used in the simulator for experimental verification.(1)The automatic control system of the offshore petroleum platform injection-production dynamic simulator is built, which is based on the Schneider island as its controller and the KingView as its PC software. The control system includes a closed-loop flow control system and data monitoring of the entrance pressure, motor frequency and so on, which lay the foundation for the application of the optimizing control policy.(2)The intelligent model-free gradient optimization control method are designed and which are applied to offshore platforms simulator. The algorithm makes the system efficiency as a target. To begin with, based on the intelligent expert thoughts, this method changes the production of the mining pump randomly and calculates the efficiency of the well in real time. Subsequently, aiming at optimizing the efficiency we determine the controlled direction of production in accordance with the efficiency changing direction. These measures can finally make the production efficiency approximate to the optimum value. The experimental results on the physical experiment system show that this optimization method can search and make the system stabilize close to the best efficiency point.(3) The GA optimization control method based on the ANN mathematical model are designed and which are applied to offshore platforms simulator. Under the premise of making a large number of conditions data acquisition, establish BP neural network model. Based on the single-well and multi-well optimization requirements, optimal system efficiency and its corresponding motor frequency could be obtained through genetic algorithm iterations. The optimization results are applied to the offshore platform simulator by the communication with the host computer. The experimental results show that the research accuracy of the algorithm is high and contrast with the above method, its reach time has obvious advantages.(4) Based on the intelligent model-free gradient algorithm and ANN mathematical model respectively, the platform integrated optimization control method are designed and which are applied to offshore platforms simulator. Based on the application of the two optimization control methods on the single-well and multi-well optimization control, integrated control program is designed further. To begin with, this method establishes the follow relationship between oil and water subsystem subsystem and carry on the efficiency optimization internally both in the oil and water subsystem subsystem. Through the commissioning tests, both the optimization control methods are better achieved the oil subsystem subsystem following water subsystem subsystem and their internal system efficiency optimization, which simultaneously solve the problem of the low operating efficiency of the platform.In conclusion, this paper take the offshore platform injection-production dynamic simulator as experimental device, the intelligent model-free gradient optimization control method and the GA optimization control method based on the ANN mathematical model are proposed in succession. Finally the platform integrated optimization control strategy is proposed. The comprehensive experimental results show that these various optimization control methods designed in this paper have more satisfactory control results, and thus has an important application value for practical engineering. |
PDF Full Text Request