Machine learning algorithms can automatically learn and extract features from a large amount of geological data to achieve fast and accurate lithology identification. In this paper, the logging data of several wells in a sandstone-type uranium deposit in Inner Mongolia were randomly divided into training sets and verification sets according to the ratio of 7∶2. The model structure was adjusted and the hyperparameters were optimized for training. BC1401, BC2802, BC4603 and BC7206 well were used for testing to realize the comparative analysis of 5 kinds of models, such as random forest, XGBoost, K value proximity algorithm, BP neural network and SMOTE-LSTM algorithm. The results show that SMOTE-LSTM model has the most superior stability and accuracy, with an accuracy of 84.6%.

The accuracy of hydraulic conductivity holds great significance for the subsequnce research and exploitation of the in-situ leaching sandstone-type uranium deposit. Taking the in-situ leaching technology of sandstone-type uranium deposit as the research object, based on the current research status of hydrogeological well completion technology and hydraulic conductivity calculation methods at home and abroad, this paper analyzes the influencing factors of hydrogeological well completion quality on hydraulic conductivity calculation. Through VB visual programming software, the loopholes of human error in calculation and curve drawing are made up, and the calculation of hydrogeological parameters of deposit is more concise, efficient and accurate.

In order to solve the problems of microwave plasma decomposition of ¹³CF₄, such as complex operation process, harsh decomposition conditions and easy corrosion of reaction tubes,the thermal catalytic decomposition of ¹³CF₄ to produce ¹³CO₂ was studied. Under the condition of anhydrous and high temperature, ¹³CF₄ was decomposed into ¹³CO₂ by gas-solid reaction between ¹³CF₄ and solid defluorination agent, using γ-Al₂O₃ as solid defluorination agent, meanwhile, the metal oxide in the defluorination agent is converted into metal fluoride. The results show that when the flow rate of reaction gas is 120 mL/min, the loading amount of catalyst is more than 60 g, the content of O₂ in feed gas mixture is more than 50%, and the temperature of high temperature tubular furnace is 900℃, the conversion of γ-Al₂${{\mathrm{O}}_{3}}_{}$ catalyst to ¹³CF₄ gas is over 90%, which has good catalytic decomposition performance and can realize continuous catalysis and decomposition. The catalytic decomposition process has the advantages of simple process, convenient operation and very high carbon yield, and can be used in industrial production.

In the in-situ leaching of uranium, uranium migration is influenced by both groundwater flow and solute diffusion, and this process can be effectively modeled using the advection-diffusion equation. Accurately modeling the variation of uranium concentration over time and space is crucial for predicting uranium migration in groundwater during in-situ leaching of uranium. Traditional numerical methods, such as the finite difference method, are computationally intensive and prone to errors while dealing with high-dimensional, complex problems. Therefore, this research aims to explore the applicability and accuracy of physics-informed neural networks (PINN) in solving the advection-diffusion equation. Through numerical simulations of the one-dimensional advection-diffusion equation, and by comparing the PINN solutions with numerical and analytical solutions. The results show that PINN provide higher accuracy and better alignment with the analytical solution over long-term simulations compared to numerical methods. Furthermore, PINN exhibit certain extrapolation capabilities. Additionally, the introduction of dropout enhances the generalization ability and convergence speed of the PINN model, confirming the potential of PINN in solving complex physical problems.

To investigate the adsorption characteristics of sodium dodecylbenzene sulfonate on scheelite surface, the water bath shaking method was used to study the adsorption capacity, kinetics, and thermodynamics. The effects of pH, reagent concentration, temperature on the adsorption characteristics were explored using kinetic and thermodynamic equation models. The results show that under pH=10, the adsorption capacity of sodium dodecylbenzene sulfonate on scheelite surface is the largest, while under pH=7, the adsorption rate is the fastest, which conforms to the pseudo-second-order kinetic equation. Within a certain range, the adsorption capacity of dodecyl benzene sulfonate on scheelite surface is positively correlated with the reagent concentration and temperature. The adsorption process conforms to the Freundlich equation, which is an endothermic reaction of heterogeneous multilayer adsorption.

In order to clarify the leaching effect of horizontal well technology on in-situ leaching of uranium, a numerical simulation method was used to study the influence of different parameters on "Horizontal Well Injection-Vertical Well Pumping" and conventional "Vertical Well Injection and Pumping". The results show that compared with "Vertical Well Injection and Pumping", the length of horizontal wells is proportional to the leaching range, the vertical leaching range of "Horizontal Well Injection-Vertical Well Pumping" is small, the plane sweep efficiency is high, the dead corner of leaching is small, and the effective utilization rate of leaching solution is high; it has a wide coverage in the horizontal direction and have significant advantages in the treatment of thin uranium ore deposits. The ratio of horizontal permeability to vertical permeability has a greater impact on the leaching of "Vertical Well Injection and Pumping". When the vertical permeability is low, the leaching effect of horizontal well sections is better than that of vertical well sections; the amount of pumping and injection fluid is positively correlated with the leaching range. As the amount of pumping and injection fluid increases, the dead corners of leaching of both "Horizontal Well Injection-Vertical Well Pumping" and "Vertical Well Injection and Pumping" decrease. The increase in the amount of pumping and injection fluid has a greater impact on the leaching effect of "Horizontal Well Injection-Vertical Well Pumping". When the flow rate reaches 210 m³/d, the leaching range of the ore layer of "Horizontal Well Injection-Vertical Well Pumping" is higher than that of generalized "Vertical Well Injection and Pumping".

A sandstone-type uranium deposit in Xinjiang belongs to auspicious prospecting stage with exploration line of 200 m×200 m, low exploration degree, thin ore bed and low uranium content. In addition, there are complex hydrogeological conditions with poor permeability (permeability coefficient is only 0.045 m/d) and high pressure head. No leaching test work has been carried out in this deposit area. In order to study the leaching of uranium resources in the deposit, field leaching test was carried out by adopting neutral leaching and improving the pressure leaching technology, the results show that the average extraction volume of single hole is 1.7 m³/h, the average mass concentration of uranium in leaching solution is 40.37 mg/L, the maximum mass concentration of uranium in single hole is 390.22 mg/L, and the annual leaching recovery is 13.39%, it provides technical support for the exploitation of uranium resources in the deposit.

The electric control valve in the ⁷⁶Ge isotope production process is controlled by the PID module of the DCS (Distributed Control System) to regulate the internal pressure of the cascade pipeline. However, when the pressure of the process system fluctuates greatly, the existing control method of electric valve can not achieve the optimal control effect, which leads to the controlled process pressure exceeding the operating safety limit and has a serious adverse impact on the safe and stable operation of the process system. Aiming at the problem of weak control performance of electric valve under abnormal pressure, this paper put forward a control scheme to optimize DCS control configuration program of electric valve. Actual operation shows that the optimized control program avoids significant fluctuations in process pressure under abnormal conditions, effectively ensuring the safe and stable operation of the ⁷⁶Ge isotope production line.

This paper focuses on the abnormal fluctuation of yield during the electron beam welding process of fuel rods. Considering that the main defects leading to the rejection of fuel rods are surface scratches and out-of-tolerance straightness, through analysis and research, it is clarified that the direct causes of the abnormal yield of fuel rods are the excessively large aperture of the welding tooling baffle, the excessive use of the welding top, and the design of the top bearing. By taking measures such as reducing the aperture of the baffle, determining the reasonable number of uses for the welding top, and designing the top bearing, defects can be controlled in a timely manner, the yield can be improved, the manufacturing cost can be further reduced, and a foundation can be laid for the subsequent exploration of fuel rod manufacturing processes.

As the mining depth of a uranium mine continues to increase, the ore body shows a trend of decreasing thickness and increasing inclination angle. This paper focuses on the steeply inclined thin ore veins of the uranium mine, utilizing the AHP-TOPSIS method for comparative analysis of different mining methods, and proposes a high-segment continuous efficient collaborative mining method suitable for the thin ore veins in this mine. Based on this, the numerical analysis is conducted using FLAC3D to examine different structural parameter schemes of the stope, analyzing the stress field, displacement field, and distribution characteristics of the plastic zone during the mining process, ultimately acquiring the optimal stope structural parameters. The results indicate that: no tensile stress was observed in the top and intermediary pillars in scheme 1 to 4, while a small amount of tensile stress occurred in schemes 5 and 6; the proportion of the plastic zone varies from 22.52% to 27.98% across scheme 1 to 6, with scheme 3 demonstrating better stability. Considering both safety and economic factors, the optimal stope structural parameters are determined to be a top pillar height of 5 m and an intermediary pillar width of 8 m (scheme 3).