This article presents a method for characterizing the system dynamics of a trapped particle in real-time and designing a controller to minimize disturbances to the particle's position. Specifically, adaptive system identification is used to determine the trap characteristics and the actuator transfer function describing the mirror voltage to trap position path. Using an internal model control scheme combined with a filtered-x least-mean-square algorithm, adaptive control was used to create a controller that minimizes a frequency weighted mean-squared-error. The dynamics associated with multiple particle sizes and materials were experimentally determined under different power levels, each case resulting in different system dynamics and demonstrating positive control results. The adaptive system identification and the controller presented automate the process of system identification and control design, enabling the automation of optical trap controller design.