Iterative Learning Algorithm with Adaptive Forgetting Function for Suppressing Compressor Torque Ripple

Torque ripple is an important concern in air conditioning compressor systems operating at low frequencies. An algorithm for suppressing compressor torque ripple is proposed to address the problems of significant torque ripple. This iterative learning algorithm with an adaptive forgetting function is applied to the speed loop, improving the learning ability under different operating frequencies. The optimal parameter configurations are achieved by analyzing control system discretization. Meanwhile, A reduced-order state observer is included to eliminate non-periodic disturbances and enhance system stability. And the output of the proposed algorithm is compensated to the given current in the q-axis The algorithm is validated on an experimental platform for air conditioning compressors, Experimental results show that the algorithm can significantly reduce the compressor torque ripple.