Abstract: In strip pickling lines, conventional fuzzy PID temperature control suffer from sensor noise, mismatched fuzzy rules, and limited adaptability. To overcome these drawbacks, a hybrid strategy for optimizing fuzzy PID that integrates local outlier factor (LOF), Kalman filter (KF) and whale optimization algorithm (WOA) is proposed. Firstly, LOF combined with a moving-average filter detects and corrects abnormal sensor readings, reducing their influence on the control loop. Secondly, KF fuses the cleaned data to suppress noise and disturbances. Finally, WOA optimises the fuzzy PID parameters on-line, minimizing manual tuning effort and improving accuracy. Simulation results show that, compared with conventional PID and standard fuzzy PID, the proposed scheme shortens settling time by 30.2 % and 17.3 %, and cuts overshoot by 2.56 % and 1.88 %, while markedly enhancing accuracy, robustness and disturbance rejection. The overall effect of temperature control during the pickling process of strip is optimized. These improvements contribute to more sustainable pickling operations, higher productivity and lower costs, and provide a transferable reference for PID enhancement in other industrial processes.