This paper proposes a novel approach to diagnosis the faults in analog circuits based on Volterra kernel and ant colony-particle swarm algorithms. In the analog circuit fault diagnosis, we use the Volterra kernel as the feature vector which makes the characteristic vector lumped Euclidean distance in serials of fault states under the same excitation signals as the fitness function. And the optimized the parameters are used to stimulate the multi- frequency sinusoidal signal. The AS-PSO hybrid algorithm is performed to find the best excitation signal parameters. Experimental results show that the proposed approach can achieve good faults diagnosis results.
