Neural network inverse control of variable frequency speed-regulating system in V/F mode

An induction motor driven by a normal and low-cost inverter running in V/F mode, named as a variable frequency speed-regulating system in V/F, is widely used, but its control performance is not good enough to meet the needs of speed-regulation. So it is useful to improve its control performance without changing the original structure of variable frequency speed-regulating system (VFSRS). Considering the induction motor and the inverter as a whole controlled object, a mathematic model of such variable frequency speed-regulating system in V/F and its inverse model, with or without compensation, are given in this paper. Constructing a neural network inverse and combining it with the variable frequency speed-regulating system in V/F, a pseudo-linear system is completed. Then a linear close-loop adjuster is designed to obtain a better speed-regulating performance. Results of experiments demonstrate that speed-regulating performances can be greatly improved using this simple method.