In the paper we design an integrated pattern recognition system. Instead of design a classifier for pattern recognition, a finite number of classifiers are simultaneously applied, and a multi-layer artificial neural network with feedback is employed to process all outputs of the individuals to get more accurate classification rate. Because of the introduction of the feedback loop, the pattern recognition system becomes a nonlinear dynamic system, not a nonlinear mapping any more. We obtain the sufficient condition on the absolute stability for the integrated network and derive a corresponding learning algorithm to ensure its stability. The system has been applied in totally unconstrained handwritten numeral recognition and its performance is excellent.

No abstract

No abstract

Under the background of the rapid development of Smart Substation in the new era, the relay protection device plays an important role in the operation of Smart Substation, which has great guarantee for the safety and stability of the power system. However, in the current situation, the relay protection equipment or the failure of the line is very common, which will not only hinder the operation of other systems, but also greatly increase the economic loss caused by system failure. With the continuous development of visual remote sensing image processing, the visualization technology is introduced into the design of remote sensing image processing system and applied in the protection system of Smart Substation. The visualized remote sensing image processing system technology can improve the efficiency of relay protection of Smart Substation. It can be seen that the application of this technology in Smart Substation has obvious superiority. Therefore, the research and improvement of the relay protection scheme for the relay protection of the new generation of Smart Substation is particularly important, which is convenient for the safe operation of the power system.