A new soft-fault diagnosis approach for analog circuits with parameter tolerance is proposed in this paper. The approach uses the fuzzy nonlinear programming (FNLP) concept to diagnose an analog circuit under test quantitatively. Node-voltage incremental equations, as constraints of FNLP equation, are built based on the sensitivity analysis. Through evaluating the parameters deviations from the solution of the FNLP equation, it enables us to state whether the actual parameters are within tolerance ranges or some components are faulty. Examples illustrate the proposed approach and show its effectiveness.
Considering the problem to diagnose incipient faults in nonlinear analog circuits, a novel approach based on fractional correlation is proposed and the application of the subband Volterra series is used in this paper. Firstly, the subband Volterra series is calculated from the input and output sequences of the circuit under test (CUT). Then the fractional correlation functions between the fault-free case and the incipient faulty cases of the CUT are derived. Using the feature vectors extracted from the fractional correlation functions, the hidden Markov model (HMM) is trained. Finally, the well-trained HMM is used to accomplish the incipient fault diagnosis. The simulations illustrate the proposed method and show its effectiveness in the incipient fault recognition capability.
While the Slope Fault Model method can solve the soft-fault diagnosis problem in linear analog circuit effectively, the challenging tolerance problem is still unsolved. In this paper, a proposed Normal Quotient Distribution approach was combined with the Slope Fault Model to handle the tolerances problem in soft-fault diagnosis for analog circuit. Firstly, the principle of the Slope Fault Model is presented, and the huge computation of traditional Slope Fault Characteristic set was reduced greatly by the elimination of superfluous features. Several typical tolerance handling methods on the ground of the Slope Fault Model were compared. Then, the approximating distribution function of the Slope Fault Characteristic was deduced and sufficient conditions were given to improve the approximation accuracy. The monotonous and continuous mapping between Normal Quotient Distribution and standard normal distribution was proved. Thus the estimation formulas about the ranges of the Slope Fault Characteristic were deduced. After that, a new test-nodes selection algorithm based on the reduced Slope Fault Characteristic ranges set was designed. Finally, two numerical experiments were done to illustrate the proposed approach and demonstrate its effectiveness.