The invention provides a new model free adaptive control algorithm. Firstly, parameters of a PDD are estimated by employing the observer technology, a controller is designed based on the obtained structure of the observer, and it is analyzed and proved that a closed-loop system is stable based on Lyapunov stability; then it is analyzed that the model free adaptive control method belongs to internal model control from structure in essence, and the algorithm is further improved based on the analysis, and an improved feedback filter is introduced so that the anti-interference of the system is improved and the robustness of the system is not reduced. According to the model free adaptive control, three problems are mainly solved: 1) the designed controller only needs measuring data inputted and outputted by the system; 2) the control method does not need a system dynamic mathematical model; and 3) the stability of the closed-loop system is analyzed based on the Lyapunov theory. Finally, simulation verification of the algorithm is performed via two simulation examples, and the simulation result indicates that the method is effective.