The existing methods of well-logging interpretation cannot be applied accurately to the exploration and evaluation of carbonate reservoirs because of the fracture development. Based on the fracture density obtained by core analysis in a carbonate reservoir located in the Ordos basin, the west of China, three types of fracture density (low, medium and high fracture density) of the target formation were identified. We investigated the effect of fractures on acoustic logging signals in time and frequency domain by Hilbert-Huang transform (HHT) and extracted 11 features in time domain and 9 features in frequency domain. Then, we reduced the features in time and frequency domain to three principal components by principal component analysis (PCA). Finally, a new prediction model of GA-SVM method based on Hilbert-Huang transform of acoustic logging data was reported to predict the fracture density. The results show that fracture density has a greater effect on the attenuation of IMF2 and IMF3 components for three different types of formation by empirical mode decomposition (EMD) analysis. The energies of Stoneley wave and shear wave have higher sensitivity than compressional wave. Compared with time domain, the distribution in high frequency domain has a greater correlation with fracture density by Hilbert spectrum and marginal spectrum. The correlation coefficients between fracture density and 9 features in frequency domain (R²=0.5~0.7) are better than the coefficients with 11 features in time domain (R²=0.3~0.5). Core analysis and interpretation of resistivity image logging support the validity and effectiveness of our model. The prediction accuracy using the features in frequency domain can reach to 82~90%, which is much higher than using the features in time domain with accuracy of 52~59%. The application with more information of acoustic logging data in our model can avoid the error in velocity picking, which points out the direction for the future prediction.

The long-term stability of bulk nanobubbles is crucial for their functional applications; however, understanding the evolution of their size distribution remains a significant challenge. While conventional characterization methods, such as Dynamic Light Scattering and Nanoparticle Tracking Analysis, provide size information, they are often sample-intensive and expensive, making them ill-suited for high-throughput or long-term dynamic monitoring of size distribution polydispersity. This research validated UV–Vis spectrophotometry as a simple, powerful tool for tracking these dynamic changes. Air nanobubbles generated via pressure oscillation-hydraulic cavitation were systematically monitored over 30 days using correlative DLS, NTA, and UV–Vis spectroscopy. A distinct two-stage evolution was identified: an initial “purification” phase marked by the dissolution of unstable bubbles, followed by a long-term “maturation” phase governed by Ostwald ripening. The Ångström exponent (n), derived from the full extinction spectrum, is a highly sensitive descriptor of this process. The evolution of n traced a unique V-shaped trajectory, which resulted in a pronounced hysteresis loop when plotted against the mean diameter from DLS. This hysteresis reveals that systems with identical mean diameters can possess vastly different distribution morphologies, which are inaccessible through traditional sizing methods alone. This research establishes full-spectrum UV–Vis analysis as a robust methodology, enabling rapid and efficient assessment of nanobubble stability and providing a deeper mechanistic understanding of their complex evolution.

The particle feeding system is a prerequisite for the realization of particle impact drilling technology. Because of the high density, the storage and flow of the steel particle are different from those of the other nonviscous particles. The differential equation of the particle movement was built with the liquid bridge force model and the discrete element method. The dynamic movement process and the distribution state of particles in the high-pressure tank were analyzed. For 1 mm steel particles, the mass flow rate decreased with the increase in water content. For 2 mm and 3 mm steel particles, the water content of 15% and 20% was the dividing point of the mass flow rate from increasing to decreasing. When the water content was 10% and 20%, the mass flow rate increased with the steel particle size. But when the water content was 30% and 40%, the mass flow rate decreased with the steel particle size. The study of the control mechanism of the uniformity and stability of particles showed that the funnel flow was the major reason causing the instability and blocking of the feeding process. This research results can provide a basis for the further improvement of the differential pressure feeding system.

In order to prevent collision of complex structure well, accurate measurement of the distance of adjacent wells must be required during the development of the unconventional reservoir. The precision of the traditional survey tools and scanning software is not accurate enough. Based on the electromagnetic anticollision measuring calculation method, the electromagnetic anticollision detection tool was designed and simulation experiments were carried out. The influence of different position relations between sensor and casing on magnetic field strength detected by the sensor was analyzed. When the sensor and the casing were parallel and the distance of them was 0.5 m∼3 m, the tool could calculate accurately the distance and orientation of the adjoining well. When the angle between the sensor and the casing was within 50°, the calculation result was accurate. This research result validated the accuracy of the principle of the electromagnetic anticollision tool while drilling cluster well, and it could provide theoretical support for the development of electromagnetic detection tools.

Turbulent drag reduction (DR) using polymers is a critical technique for energy conserva-tion in fluid transport systems. Traditional monitoring methods relying on pressure transducers are intrusive and lack real-time turbulence characterization. This study pro-poses a novel non-intrusive intelligent monitoring system based on Fiber Bragg Grating (FBG) sensing and Artificial Intelligence (AI). An experimental setup was constructed to investigate the DR performance of polymer solutions. FBG sensors were utilized to capture the optical spectrum shift induced by turbulent flow fluctuations. A deep learning model was trained to correlate the optical signal features with the drag reduction rate. Results demonstrated that the AI model achieved high prediction accuracy (R² > 0.95), effectively replacing complex fluid dynamic calculations with optical signal analysis. This work provides a promising approach for real-time, non-intrusive monitoring of fluid flow characteristics in industrial applications.