Background: Aiming at the three-phase voltage source PWM rectifier (VSR), there have been some researches on the neural network control strategies, but the unbalanced power grid condition is always neglected. Meanwhile, about the self-detection technology of the unbalanced power grid voltage, there are still few research.
Methods: Under the unbalanced power grid conditions, this work presents a power grid voltage sensorless control strategy of the three-phase VSR. Based on the radial basis function neural network(RBF) theory, a newly PI controller with parameter self-tuning function is studied, and the RBF-PI controller is used for the closed-loop controls of the voltage and current variables. By means of T/4 delay method, the positive- and negative-sequence components of the power grid side current, and those of the equivalent virtual flux-linkage of unbalanced power grid are extracted. From the equivalent virtual flux-linkage, the unbalanced power grid voltage is online reconstructed.
Result: Under the load mutation condition, the power grid voltage sensorless control of three-phase VSR is achieved, the system response characteristics are analyzed. From the simulation experimental results, it can be seen that the unbalanced power grid voltage can be quickly tracked, and the sinusoidal control of the power grid side current can be achieved. In addition, the control system has a series of advantages, such as a faster dynamic response speed, a stronger robustness and a smaller DC side voltage ripple.
Conclusion: The proposed unbalanced power grid voltage sensorless control strategy of three-phase VSR is effective.