Abstarct: Induction machines are one of the most widely used machines in industry due to their simple structure, high efficiency and low maintenance cost. However, failure of these machines, especially in the rotor section, can lead to costly downtime in industrial processes. In this thesis, a practical method for monitoring the condition and detecting rotor unbalance fault in low-power induction motors is presented baxsed on stray magnetic flux measurement. Unlike many conventional methods that require current measurement or sensor installation inside the motor, the proposed method uses two coils with a phase difference of 90 degrees, which are installed outside the motor. This method allows drawing an orbit by analyzing the flux signal in both horizontal and vertical directions, and changes in the resulting ellipse shape will be an indication of کلمات کلیدی: خطای روتور، شار سرگردان، تحلیل مداری، روش عملگر انرژی، شاخص‌های ‌خطا، پایش وضعیت، روش یکسوکننده شار سرگردان Abstract Induction machines are one of the most widely used machines in industry due to their simple structure, high efficiency and low maintenance cost. However, failure of these machines, especially in the rotor section, can lead to costly downtime in industrial processes. In this thesis, a practical method for monitoring the condition and detecting rotor unbalance fault in low-power induction motors is presented baxsed on stray magnetic flux measurement. Unlike many conventional methods that require current measurement or sensor installation inside the motor, the proposed method uses two coils with a phase difference of 90 degrees, which are installed outside the motor. This method allows drawing an orbit by analyzing the flux signal in both horizontal and vertical directions, and changes in the resulting ellipse shape will be an indication of the presence and severity of the fault. Also, by using frequency indices and energy operator analyses (TKEO) and the stray flux rectifier method, the sensitivity of the system to different fault intensities at different speeds has been investigated. Comparison of the results with the spectral analysis method (FFT) shows that the proposed method, in addition to reducing the effect of spectral leakage, also performs better in detecting early faults. Since there is no need to install sensors in specific locations or directly access the inside of the motor, this method has a significant advantage over conventional methods in terms of implementation cost, ease of installation, and accuracy in detection.