In the process industry, actuators regulate the flow of process masses in response to instructions from a controller, which achieves the control objectives of the system. However, many factors cause the performance of the actuator to degrade. Therefore, it is of great significance to evaluate the actuator operating performance correctly for equipment maintenance and system optimization. In response, this paper designs an improved multifractal method based on statistical moment functions. And a set of indicators-based degradation assessment system is constructed. First, the time series collected from the actuator is divided into two different density zones, where evaluation indicators are established respectively to qualitatively describe the performance from different perspectives. Then, a degradation index calculated by weighted fusion is applied to quantitatively evaluate the dynamic degradation state of the actuator. Finally, the effectiveness and practicability of the method are experimentally verified through two field cases in thermal power units.