Considering the load characteristics of the solid-state induction heating power supply, the resonant inverter load matching schemes are compared and analyzed in this paper. Especially, having a detailed research on the load matching schemes with passive component, and proposing the design processes of load matching, which proved by experiments. After analyzing the merits and drawbacks of neural network and D-S Evidence theory respectively, a method of hierarchical information fusion condition evaluation based on neural network and D-S Evidential theory is introduced in this paper, the principle and steps are discussed in details. We choose Windows 2000 as the operating system, SQL Server 2000 as backstage supporter's database, Delphi as the front developing instrument, to set up the resonant inverter load matching scheme analysis system. After doing more than 80 experiments on 30 different induction heating power supplies, the system have good behavior over fast speed, excellent stability, high veracity, and high reliability.
|