To address the challenges of low signal-to-noise ratio （SNR） and the lack of clean ground truth labels for distributed acoustic sensing （DAS） and distributed temperature sensing （DTS） data during hydraulic fracturing，a physics-informed self-supervised blind denoising method is proposed. Instead of relying on clean reference data，this method constructs self-supervised training samples using a checkerboard alternating sampling strategy on time-depth sub-blocks and enables the denoising network to learn noise distributions adaptively. Furthermore，physical constraints such as thermal conduction smoothness，edge preservation，and wavefield coherence are introduced for DTS and DAS characteristics，respectively. A spatially weighted loss function based on perforation cluster positions is also established to prioritize the preservation of key response features near clusters. It is demonstrated that the proposed method outperforms traditional filtering methods in terms of SNR and structural similarity （SSIM）. The denoised data significantly enhances the clarity of cooling fronts and acoustic energy bands at fluid entry points，which effectively improves the accuracy of fracturing event recognition and cluster efficiency evaluation.

The deep shale oil reservoir of the Fengcheng Formation in Mabei area features complex geological conditions and its fracturing stimulation is faced with multiple challenges，including high in-situ stress，strong heterogeneity，and densely laminated rock texture. Taking the C2-C4 layers as the research object，this study systematically investigates the mineral composition，laminar texture，rock mechanical properties，and in-situ stress distribution of the reservoir to reveal their control mechanisms on hydraulic fracture initiation and propagation，as well as proppant transport. The results indicate that the reservoir is predominantly composed of high-brittleness minerals，with the combined content of quartz and carbonate rocks exceeding 70%. However，the mineral composition varies dramatically at the centimeter to millimeter scale，forming dense rigid-brittle laminae and lithological interlayers which significantly enhance mechanical anisotropy and interface strength of rocks. The overall in-situ stress is high，with the minimum horizontal principal stress exceeding 80 MPa，while the horizontal stress difference is only 7.0-9.8 MPa. Theoretically，this low stress difference favors the formation of complex fracture networks. Yet，due to the dual effects of high stress magnitudes and laminar interfaces，vertical fracture propagation suffers from step-like restrictions，with fracture heights generally less than 20 m，and thus，it is difficult to achieve thorough stimulation of multiple layers. Natural fractures intersecting hydraulic fractures at small intersection angles may induce branch fractures to increase fracture network complexity；at high intersection angles，they primarily act as conduits for fluid filtration loss. Under the “composite layer effect，” the effectiveness of layer penetration of hydraulic fractures is low，with width reduction at points of fracture deflection exceeding 60%. Proppant can only pass through fractures with widths 2.7 times larger than proppant particle sizes，and accordingly，proppants of conventional sizes are found hard to enter secondary fractures. This study reveals the control mechanism of the “composite layer effect” on hydraulic fracture propagation and proppant transport in the Mabei Fengcheng Formation，provides a theoretical basis for volume fracturing parameter optimization and fracture network regulation in deep complex shale oil reservoirs，and offers direct technical support for the efficient development of shale oil in Mabei area.