<p>The study area is rich in shale gas resources and has reached the stage of comprehensive development. Shale gas extraction poses risks such as induced seismicity and well closure, compounded by the limited availability of fixed seismic monitoring stations nearby. To address these challenges, a dense observation array was developed within the study area to monitor and analyze microseismic activity during hydraulic fracturing. Microseismic events generated by hydraulic fracturing typically exhibit low amplitude and signal-to-noise ratio, rendering traditional manual analysis methods impractical. To overcome these limitations, an innovative artificial intelligence method combining picking-association-location (PAL) and match-expand-shift-stack (MESS) techniques (PALM) has been utilized for automated seismic detection. Numerous factors influence the accuracy of microseismic detection and localization. To evaluate these factors, the effects of various velocity structure models, instrument types, and station distributions on seismic location were analyzed and compared. The results indicate that the PALM method significantly mitigates the influence of velocity structure models on seismic location accuracy. Additionally, the use of broadband seismic instruments and a uniform station distribution enhances the precision of seismic location results. Furthermore, by integrating data from different types of observation instruments, a comprehensive seismic catalog for the study area was established. These findings not only enhance seismic location accuracy but also provide valuable guidance for optimizing regional seismic monitoring network design and improving seismic risk assessment.</p>