A method for multiple soft fault diagnosis of linear analog circuits

The problem of testing and diagnosing multiple soft faults in analog linear electronic circuits is discussed in this paper. An efficient method which detects faulty elements and evaluates their parameters is developed. The algorithm employs two diagnostic tests, carried out in AC state, providing rms values of the measured voltages. One of them, called a principal test, is used by the diagnosis procedure whereas the other test serves the validation purpose. Each of the voltages is an algebraic function of the tested parameters. The proposed algorithm solves least squares problem by minimizing the sum of squares of the differences between the measured voltages and the parametrized functions using the Levenberg-Marquardt method. The algorithm is applied to all testable sets of the elements and every time the result is verified using the validating test. The method may provide more than one set of the parameters which satisfy both tests due to ambiguity groups existing in the circuit. To adapt the method to the conditions encountered in practice, perturbations of the fault-free parameters and measurement uncertainty are taken into account. For illustration of the method the diagnoses of three benchmark linear electronic circuits are discussed using laboratory performed tests.