Optimizing And Achieving Of The Analog Circuits Diagnosis Stimulus Based On Volterra Kernel

For analog circuit diagnosis based on Volterra kernel, parameters of testwaveform determine the difference of fault characteristics, and directly affect theaccuracy and efficiency of the diagnosis. At present, the measurement of Volterrakernel and parameter optimization of test waveform still need to be further perfectand improved. On the base of consulting a great amount of materials about circuitfault diagnosis, then the design method of optimizing the stimulus of analog circuitsbased on Volterra kernel is presented.Volterra kernel measurement is a necessary link, when it as a diagnostic featureof the circuit faults diagnosis system. As we know, for the nonlinear circuit, themeasurement of the Volterra kernel is troublesome in calculation. Based on theanalysis of the frequency domain, this article introduces the technology of frequencydomain separation and the method for selecting test frequency. And a way tomeasuring the Volterra kernel using the signal from multiple frequency is describedin this paper. On the problem of optimizing stimulus, the particle swarm optimizationalgorithm is briefly introduced and a perfected adaptive particle swarm algorithmwith dynamically changing inertia weight is introduced due to the classical PSO withsome deficiencies. This paper gives mixed ant colony particle swarm algorithm andtest stimulus optimization using local optimal solution characteristics of the antcolony algorithm. Besides theoretical research, the paper designs hardware andsoftware of the test stimulus optimization system and builds the test stimulusoptimization system. Hardware circuit is made up of four parts: a main control unit,signal DDS technology unit, data acquisition unit and communication interface.Software design can be divided into the upper computer part and the lower computerpart.After finished the design of the peripheral hardware circuit of single-chip anddesigned of the corresponding software. The optimizing system passed the experiment on the typical nonlinear analog circuit for many times, through theexperimental results and waveforms to confirm the feasibility of this design.