Abstract : To overcome the dependence on motor parameters in speed estimation, some scholars have proposed extracting speed using frequency components related to rotational speed in tooth harmonic signals. Therefore, speed estimation involves extracting relevant frequencies from the tooth harmonic signal and calculating the speed based on their relationship with rotational speed. Since rotational speed is time-varying and closed-loop speed control is required, the detection algorithm must have low computational complexity and be able to track time-varying signals. Especially at low speeds, measurement sensitivity and anti-interference issues need to be addressed; these are key technologies that urgently need to be solved in current sensorless vector control technology. This paper mainly studies a fast harmonic detection algorithm based on recursive least squares to meet the requirements of speed identification for both speed and sensitivity. Keywords : Sensorless inverter, tooth harmonics, digital signal processing, least squares algorithm. The basic control idea of ​​sensorless vector control is to perform flux observation and speed observation. This paper focuses on speed estimation using the stator current tooth harmonic method. This method mainly relies on digital signal processing technology for harmonic analysis and harmonic signal frequency extraction. The shortcomings of modern spectrum estimation techniques widely used today, such as FFT, autocorrelation power spectrum estimation, and AR models, are that to ensure estimation accuracy, the required sampling time is often relatively long, the real-time capability is relatively poor, and they are highly susceptible to noise interference, resulting in large estimation errors at low speeds. In short, the main problems to be solved by this method are anti-interference at low speeds, measurement sensitivity, and real-time processing capability. Therefore, to truly put this method into practical use, efforts are needed in both theoretical and technical aspects. The purpose of this paper is to solve the problems of measurement sensitivity and anti-interference at low speeds. Solving the problems existing under low-speed operation is of great significance for the practical application of this method. 1. Related Algorithms for Speed ​​Detection Using Tooth Harmonics If the motor is under load, the rotor current and rotor tooth electromotive force harmonics will also interact with each other, along with the basic components of the air gap flux. [b][align=center]For details, please click: Speed ​​Estimation of Sensorless Controlled Frequency Converter Based on Harmonic Detection[/align][/b]