Fault diagnosis of gas-turbine power units with the derivative-free nonlinear Kalman Filter

摘要: Abstract A method is developed for diagnosing faults and cyberattacks in electric power generation units that consist of a gas-turbine synchronous generator. By proving such unit differentially flat its transformation into an input–output linearized form becomes possible. Moreover, by applying the Derivative-free nonlinear Kalman Filter, state estimation performed. The latter filtering method, consists Filter's recursion on equivalent model unit, as well inverse providing estimates initial system. subtracting estimated outputs Filter from measured residuals’ sequence generated. residuals undergo statistical processing. It shown sum squares vectors, weighted associated covariance matrix, forms stochastic variable follows χ2 distribution. exploiting properties this distribution, confidence intervals are defined, which allow detecting unit's malfunctioning. As long aforementioned remains within previous normal functioning inferred. Otherwise, fault or cyber-attack detected. also subspaces system's state-space model, isolation can be