A method for estimating the performance of and/or detecting faults in a multi-component system, eg a gas turbine (fig 4). The system can be defined by a function h() of the performance parameters x and operating parameters w defining an operating condition of the system. An initial population of strings is created consisting of a number of elements corresponding to estimated values of performance and operating parameters. The method comprising optimising an objective function J(x,w) which relates the function h() to measured parameters Z based on MSBbias/SB which is the maximum number of fault effected measurements using genetic algorithms to create successive populations of strings. A selection process is used to generate the next population of strings by eliminating strings from the population which have a low fitness value based on the objective function and creates cross-over and mutation strings to add to the population. Multiple calculations of the objective function are performed with a different MSBbias/SB eliminated then selecting the formula used to produce the minimum calculation of the optimisation function.
展开▼