During research on SMA wires the prototype of linear position actuator was built. The shape memory alloy (SMA) wires used in construction of the actuator are nonlinear and time variant. Thus, it was decided to use a fuzzy controller to control the actuator. However, because of the nature of SMA wires which work by changes of their temperature, after a few minutes of continuous work the actuator did not work accurately. In other words, the existing singleton values in a knowledge base were inappropriate. That is why each time after longer continuous work of the actuator it was needed to manually find and change the mentioned values. It took a lot of time and effort so eventually it was decided to create a real time auto-tuning algorithm which could identify crucial parameters of the actuator each time when it was needed. Together with initial values learning algorithm the advanced controlling of SMA actuator was created.