State estimation method for deepwater drilling riser system based on monitoring information

Abstract

The state of the deepwater drilling riser system is critical for risk management. Monitoring is the most direct and accurate way to obtain the state. However, the large-scale deepwater drilling riser system makes it difficult to deploy global monitoring instruments, resulting in only local states that can be obtained by monitoring. Obtaining the global state of the deepwater drilling riser system has long been an engineering challenge. A state estimation methodology based on the locally monitored state and prior model of the drilling riser is proposed to solve the challenge. The locally monitored state includes the current state, floating drilling platform motion state, and riser string local acceleration state. The prior model includes the structural mechanics model of the drilling riser and various load models. A backpropagation model based on cross entropy and gradient descent method is established to modify the local prior model according to the locally monitored state. A global state estimation model is established by extending the modified local prior model with the proposed construction strategy. A drilling riser system with a water depth of 1000 m is selected to test the state estimation method. The results show that the proposed state estimation method can effectively estimate the global state of the deepwater drilling riser system. The estimation effect of the drilling riser state caused by the wave is relatively better than that caused by the current. The monitoring location has a significant effect on the estimation results of the drilling riser state caused by the current.