Abstract
Marine gas turbines have been widely used and developed in the field of marine power. It is important to make them operated safely and efficiently. In this paper, a marine triaxial gas turbine is taken as an example to study the method of estimating the health state of the gas path using extended Kalman filter (EKF). To verify the accuracy of EKF, a comparison was made between linearized Kalman filtering (LKF) and EKF. In addition, the sequential quadratic programming (SQP) algorithm is used to seek the performance in case of gas path abnormal. The combination of parameter estimation and performance seeking forms a comprehensive method for diagnosis and optimization of marine gas turbines.
The results show that the EKF method is an effective method for combining nonlinear systems with traditional Kalman filter. EKF has a good estimation effect on the gas path health state under different operating conditions. Also, the marine triaxial gas turbine achieved the target performance under the constraints of the SQP algorithm. Performance seeking restores the output power of the marine gas turbine and reduces the inlet and outlet temperatures of turbines. It can effectively prevent the problem of excessive combustion and ensure the safe and stable operation of the marine gas turbine.
State and Parameter Estimation in Hall-Héroult Cells using Iterated Extended Kalman Filter
