The invention discloses a lumped disturbance parameter observation and
permanent magnet synchronous motor control method and
system, and the method comprises the steps: collecting the phase current of a motor in each
control period, designing a two-degree-of-freedom filter to filter
DC bias and high-
frequency noise signals, i.e., setting two low-pass filters which are connected in parallel for subtraction, independent suppression of a high-
frequency noise signal and direct-current bias is realized by adjusting
cut-off frequencies of the two, and decoupling of a filtering effect is realized; then, according to the filtered current
signal, a super-spiral sliding-mode observer containing an equivalent feedback
gain and an exponential term is constructed to observe a lumped disturbance parameter, the equivalent feedback
gain is in an integral form, high-
frequency noise can be effectively suppressed, the dynamic convergence process of the observer is smoother due to the additionally designed exponential term, and the lumped disturbance parameter is obtained. And the lumped disturbance parameter observation value is used for model-free current prediction of the motor. According to the invention, direct-current bias and high-frequency
noise in the current sampling
signal can be effectively filtered out, and the observation precision and control precision of the super-spiral sliding-mode observer are improved.