The design of robust MIMO neural network disturbance observer for multi-variable system

1. School of Automation, Southeast University, Nanjing 210096,Key Laboratory of Measurement and Control of Complex Systems of Engineering, Ministry of Education, P.R.ChinaE-mail: lsh@seu.edu.cn2. School of Hydraulic, Energy and Power Engineering, Yangzhou University, Yangzhou 225127E-mail: juanli@yzu.edu.cnAbstract: Multi-variable systems widely exist in the practical engineering control systems whose performances are alwaysseverely interrupted by strong disturbances including unmodeled dynamics, parameter variations, couplings and external dis-turbances. Disturbance observer (DOB) is known as an effective technique to estimate disturbances and has been extensivelyapplied for feed-forward compensation design in the presence of disturbances. Yet many disturbance observer techniques inprevious literature are just used for single-input-single-output (SISO) systems or the DOBs can be applied in the multi-variablesystems, but the DOBs are still SISO DOBs. A decoupled robust multi-input-multi-output neural network disturbance observer(MNNDOB) is designed for the multi-input-multi-output (MIMO) systems. Simulation results on the mixing tank show that theproposed method has better disturbance estimation performance when there are severe model mismatches compared with theMIMO linear disturbance observer.Key Words: Multi-variable system, Model mismatches, MIMO neural network disturbance observer, Disturbance estimation