| Two-stage matrix converter (TSMC) inherits many ideal characteristics from conventional matrix converter (CMC), and has other merits, such as: reduced number of switches, simple and reliable inversion. So TSMC is a novel matrix converter with huge development potentials. But TSMC has no energy storage device in direct current stage, all kinds of no-ideal operation conditions and disturbances such as voltage source's sudden change and load disturbance will deteriorate its output performances. In order to conquer the disturbances, this paper researched the closed-loop control strategy of output voltage.TSMC is a nonlinear system when runs in switch-state, it's difficult to set up its precise math model. Due to its simple structure, convenience of adjusting parameters, good stability, the conventional PID controller is widely applied in process control and movement control, but it adapts not well to time-varying, nonlinear controlled device. Therefore, it's necessary to introduce intelligence control technology which doesn't rely on its controlled device's math model and is widely applied in nonlinear control system. Single neuron has self-learning and self-adjusting ability, and doesn't rely on its controlled device's math model in process control. This paper proposed a strategy of closed-loop control of TSMC based on single neuron PID controller. The back-feed and reference voltage are transformed into direct current voltage by 3s/2r transformation, two single neuron PID controllers are used to track the reference of output voltage. Meanwhile, experts control or fuzzy control is used to adjust the gain of the single neuron PID controller, furthermore, to improve the performance of controller. Compared with PI controllers through simulation, the simulation result shows the closed-loop system with single neuron PID controllers has good dynamic,steady performance and strong robustness. At last, a TSMC experiment platform was set up, circuits and program were debugged.
|
PDF Full Text Request