The invention relates to the technical field of distribution network low-
voltage problem treatment, and particularly discloses a distribution network low-
voltage problem treatment method and
system based on
big data, and the method comprises the steps: collecting
voltage waveform data, current waveform data and load impedance data of a three-phase circuit in real time through a multi-source sensing device, and generating an original electrical parameter matrix; carrying out inter-
phase difference feature extraction on the original electrical parameter matrix to obtain the
phase offset, the impedance deviation coefficient and the
power factor angle difference value of each phase; inputting the
phase offset into a trained
convolutional neural network for
phase compensation amount prediction, and generating a
phase reconstruction parameter set; a multivariable PI D controller containing impedance feedforward is constructed,
phase reconstruction parameters and real-time impedance deviation are input, proportional
gain and
phase adjustment integral time are corrected through impedance, and cooperative
control parameters of an overload phase are generated, so that the
overall efficiency and stability of a power
system can be remarkably improved, energy waste is reduced, the operation cost is reduced, and the reliability of the power
system is improved. And a technical basis is provided for intelligent upgrading of a power system.