Multi-segment fuzzy control for start-up optimizing of LCC-based high-voltage power supply

Abstract

Widely applied in medical, industrial, and military fields, the high-voltage power supply based on LCC resonant converter (LCC) is ought to establish tens kilovolts of output voltage within milliseconds according to many standards. And any overshoot during this start-up process is not allowed because of the issue or even deadly consequence it may lead. Traditional linear control schemes are powerless to fulfill these two conflicting requirements, especially for the LCC with nonlinear gain characteristics. In this paper, a multi-segment fuzzy control is proposed to tackle this problem. Firstly, the nonlinear gain curve of LCC is piecewise linearized based on its large-signal model, and the control requirements for different segments are determined. Then the control parameters for each part are selected by an off-line fuzzy algorithm. The allowed minimum frequency during the start-up is also piecewise linearized so that the gain ability of LCC is utilized. Finally, the proposed control scheme is verified with an 80kW/150kV X-ray power supply. A low-cost controller is used to implement the start-up control scheme since the control parameters can be pre-calculated and stored in the controller. Compared with the commonly used linear PI control, the non-overshoot start-up time with the proposed control scheme is reduced by more than 60%.

Keywords

• LCC resonant converter
• High-voltage power supply
• Optimized start-up
• Fuzzy algorithm
• Multi-segment