To enhance the bandwidth and real-time capabilities of Radio Frequency (RF) spectrum analysis, the optical real-time Fourier transform method has been proposed. The optical real-time Fourier transform method based on the temporal Talbot effect has the advantage of a simple structure by using optical pulse sampling and dispersion delay structure. However, its frequency measurement bandwidth is limited by the optical pulse's repetition rate. Addressing this limitation, a real-time spectrum analysis scheme of wideband RF signals based on the fractional temporal Talbot effect is proposed and demonstrated. Based on the sampling and dispersion structure, the scheme realizes the mapping of the RF signal frequency to the optical pulse time interval. At the same time, the repetition rate of the optical pulse before sampling is multiplied by passing through a dispersive element satisfying the fractional Talbot distance in advance. The frequency measurement bandwidth of the system can be significantly improved by using the fractional Talbot effect. A proof-of-concept experiment is carried out to test the performance of the proposed scheme. The repetition period of the optical pulse is set to 151.5 ps, that is, the repetition frequency is 6.6 GHz. Each pulse has a Gaussian shape and the full width at half maximum of the pulse is approximately 30 ps. Dispersion compensation fiber is used to provide dispersion for the system. The total dispersion value of the two sections of dispersion is about 3 650 ps², which is about 0.1% different from the theoretical result. The RF signal to be measured is generated by a RF signal generator. The output optical signal is converted into an electrical signal by a 40 GHz bandwidth photodetector and recorded by a sampling oscilloscope with a bandwidth of 50 GHz. Comparing the experimental results under integer-order, 3rd-order fractional, and 9th-order fractional temporal Talbot conditions, it is verified that the measurement frequency bandwidth of the system increases with the order of the temporal Talbot effect. Real-time spectral analysis of single-tone and two-tone RF signals within a 29.7 GHz bandwidth is achieved using the 9th-order fractional temporal Talbot effect. Numerical simulation is carried out to achieve time-frequency analysis of a large-bandwidth linear chirp signal. Based on the 3rd-order fractional temporal Talbot effect, a linear chirp signal with a frequency range of 2~13 GHz and a chirp rate of 2.2 GHz/ns is successfully identified. Numerical simulation results further verify that this scheme can effectively analyze frequency transient signals. The main causes of frequency measurement errors include the time jitter of the input optical pulse train, the limited bandwidth of the pulse detection system, the deviation between the dispersion value and the theoretical value, high-order dispersion terms, etc. In the experiment, the time jitter root mean square value of the optical pulse is approximately 1 ps, which is 0.066% of the period of the optical pulse train. When the frequency measurement bandwidth is 29.7 GHz, the frequency measurement error caused by the time jitter is about 200 MHz. In order to improve frequency measurement accuracy, methods such as reducing optical pulse jitter, increasing the bandwidth of the pulse detection system, and compensating for high-order dispersion can be used. It should be noted that the increase of frequency measurement bandwidth will sacrifice the frequency resolution of the system. In practical applications, the requirements of system bandwidth and frequency resolution should be fully considered to select the order of the fractional temporal Talbot effect. With its advantages of simple structure, large bandwidth, and real-time processing, this scheme has potential application value in the fields of broadband radar, cognitive radio, and other fields.