In the microfluidic system, the utilization of superparamagnetic particles to achieve the capture, separation and detection of targets has broad application prospect in the field of biological analysis. However, the rapid and accurate automatic transmission of particles is a major challenge for the on-chip biological analysis. To solve this problem, this paper studies the modeling and motion control of particles in the microfluidic system. First, an electromagnetic coil with a tapered tip iron core is designed, and a model of the magnetic field distribution of the electromagnetic coil is formulated. Then, a dynamic model of the particles in the microfluidic chip is established, which uses a high-gain extended state observer. To estimate and suppress the uncertainty of model parameters and environmental interference, a closed-loop motion control system is designed. Finally, it is evaluated through experiments. Experimental results show that the model and control method proposed in this paper can realize the identification, tracking and motion control of superparamagnetic particles in microfluidic channels in different environments. The maximum average error of motion is 3. 37 μm, and the maximum root mean square error is 4. 29 μm.