The study of investment networks is crucial to the promotion of regional integration. Based on the data of enterprises’ outward investment from 2005 to 2022, the nodes, patterns and structural evolution characteristics of the investment network as a whole and in different industries were analyzed using social network analysis. The investment linkage matrix was also constructed and the SLM(spatial lag model) was used to explore the externalities of the investment network. The results show that the power polarization of investment network nodes is obvious, with Chengdu and Chongqing occupying an absolutely dominant position. The investment network power of manufacturing and financial industries shows ring and fragmented distribution, respectively. The investment network pattern evolves towards multi-core equilibrium, and the manufacturing network pattern is more mature. The investment network has an obvious core-edge structure, but the difference in core degree gradually decreases. The investment network has significant positive externalities on economic growth, showing a trend of decreasing and then increasing. The manufacturing investment network has a stronger role in promoting economic growth. Consumer markets and government policies are important factors driving economic growth in local and linked cities.

Production optimization is a key component of closed-loop reservoir management, aimed at maximizing economic benefits by fully exploiting reservoir potential through the identification of optimal injection and production parameter schemes. The use of numerical simulators during the search for the optimal solution significantly increases time costs, making it challenging to rapidly determine the best scheme, despite its important engineering implications. An objective function for production optimization in waterflood reservoirs was constructed based on factors such as the reservoir numerical simulation model, optimized time steps, production and injection well control strategies, reservoir responses, and cost and profit parameters. Research was conducted on the method of using surrogate models to replace numerical simulators in production optimization, based on sampling criteria for the surrogate model. In order to address the lack of physical significance in surrogate models and the difficulty in achieving the same precision as numerical simulators during optimization, a relationship for an ensemble surrogate model was established. Validation results on two reservoir models, three-channel and Egg, show that the proposed ensemble surrogate model optimization method improves economic benefits by 4.81% and 4.75% respectively, compared to conventional optimization methods across all tested reservoir models. In order to verify the application effect of the proposed method in the actual oilfield field management scenario, the Bruges oilfield is taken as an example to carry out the test. The proposed integrated agent model optimization method can increase the economic benefit by 2.45% compared with the conventional optimization method. This proves that the proposed method can achieve satisfactory results in solving the problem of injection and production optimization in large-scale reservoirs.

Loose sandstone oil and gas fields often experience sand production during development due to factors such as declining formation pressure and increasing water content, which can have a serious impact on oil recovery and production, and even lead to production shutdowns. Taking the Linhe Formation, the main reservoir of Xinghua X Block in Bayan Oilfield, as the research object, based on the classic model of sand production prediction, the TAW-1000 rock mechanics test equipment and sand control screen comprehensive simulation test system were used to carry out the sand production trend analysis and sand control parameter optimization experiments for the target reservoir. Based on model analysis and experimental evaluation, the sand production probability of the reservoir in the early stage of development of Bayan Oilfield block was predicted, and the law of reservoir rock strength changing with saturation rate was clarified. The sand control technology and screen accuracy parameters suitable for the target reservoir were optimized and determined. The study can provide important reference for the sand control technology measures and intervention timing adopted for the development of X block in the target oilfield, and has certain production guidance significance.

To improve the reliability of security monitoring results and achieve intelligent monitoring of buildings, research and design of building security intelligent monitoring methods were carried out based on sensor data fusion and scale invariant feature transform(SIFT) feature matching. Cameras, infrared detectors, vibration sensors were installed, building security data was collected, weighted average method was introduced to fuse and compress preprocessed data with point cloud data. Color images was converted into grayscale images, key points (feature points) and corresponding descriptors were extracted, and automatic stitching of building security monitoring images was designed based on SIFT feature matching. In the spliced video frames, the inter frame difference method was used to identify moving targets, and a threshold was set to distinguish between foreground(moving targets) and background, achieving tracking and intelligent monitoring of video moving targets. The comparative experimental results show that the designed method has a good effect in practical applications. This method can accurately identify all characters appearing in the monitoring interface, meeting the requirements of intelligent monitoring for building security.

A flexible sorting system for industrial intelligent robots based on machine vision has been designed, focusing on identifying and sorting plastic shell samples of remote controllers. The system uses a CMOS camera and adaptive lens, combined with a white light strip source, to ensure high-precision image capture. Through camera calibration and hand eye calibration, the robot has achieved precise positioning of the workpiece position. The system software design includes visual software settings and industrial robot software design, realizing the automation process of workpiece recognition, positioning, grasping, and placement. The results show that the system can efficiently and accurately complete sorting tasks, with high automation, intelligence, and flexibility.

The traditional urban lighting inspection method has many shortcomings, and the unmanned aerial vehicle inspection solution has been put into use in the fields of electricity, water conservancy, and urban monitoring and has achieved good results. Research on unmanned aerial vehicle intelligent inspection technology in the field of urban lighting, introducing technologies such as deep learning based urban road surface illumination collection and visual recognition based municipal infrastructure disease monitoring to achieve efficient and accurate inspection of urban road lighting facilities. In the data processing stage, fault detection and recognition are performed using a trained YOLOv5 deep convolutional neural network (CNN) model, combined with techniques such as adaptive anchor box computation and Mosaic data augmentation. The experimental results show that the system can effectively detect various types of defects in urban lighting and complement traditional inspection methods, accurately detecting potential faults in urban lighting facilities and providing data support for subsequent maintenance. This system provides an effective solution for the digital and intelligent management of future urban lighting facilities, and has great application prospects and promotion value.

The First Affiliated Hospital of Henan University of Science and Technology’s Oncology Hospital utilizes a high-difference monolithic main and podium raft foundation. To ensure construction safety, it is imperative to control and analyze the settlement differential of this foundation. The Midas GTS/NX finite element software was employed to construct a load-bearing structural model of the foundation pit, assessing the impact of the superstructure’s self-weight and the variation in construction loads on the differential settlement of the foundation. The findings demonstrate that with the increment of the superstructure’s load, the monolithic raft foundation exhibits continuous settlement at the high-difference junction without abrupt changes,the settlement of the main and podium raft foundation increases nonlinearly with the augmentation of floor loads, employing a monolithic raft foundation can effectively manage the settlement differential of the high-difference foundation.

The influence of landmark buildings in Lanzhou was mainly analyzed and discussed, aiming to study the relationship between landmark buildings’ influence, traffic attributes and landscape attributes. Urban vitality was taken as the influence index of landmark buildings, traffic accessibility represents the traffic attribute of landmark buildings, population thermal data represents the landscape attractiveness of landmark buildings, and the driving factors of landmark buildings’ influence were explored through correlation analysis. The results show that the traffic attributes and landscape attributes of landmark buildings are positively correlated with the impact of landmark buildings. The promotion effect of traffic accessibility and landscape attractiveness on the influence of landmark buildings depends on the road traffic level of the location of landmark buildings. When accessibility is poor, the impact of traffic accessibility on landmark buildings is more important than the landscape attraction of landmark buildings themselves. When the degree of transportation convenience reaches a certain level, the promoting effect of transportation accessibility on the influence of landmark buildings will decline. Improving the transportation accessibility and landscape attractiveness of landmark buildings can both enhance the urban vitality around landmark buildings and thus enhance the influence of landmark buildings. Based on research results, the corresponding planning and design suggestions are put forward.

In order to study the creep effect of concrete-filled steel tube tied arch bridge, taking 1-128 m concrete-filled steel tube arch bridge as an example, a three-dimensional finite element calculation model was established. According to the creep model of concrete-filled steel tube under the action of steel content, core concrete strength and temperature proposed by Jung Ling Chang, the influence of creep on the long-term deformation and internal force of concrete-filled steel tube tied arch bridge was analyzed. The results show that the arch rib creep will cause the arch rib and tie beam of the concrete-filled steel tube tie arch bridge to produce flexural deformation, and the arch rib stress will redistribute, resulting in the arch rib stress increasing and the core concrete stress decreasing. The stress redistribution of arch rib, girder and arch rib caused by the creep of arch rib under the consideration of temperature is greater than that without the consideration of temperature. The influence of temperature on creep should be considered in bridge design, construction monitoring and later use stage, so as to reasonably predict long-term creep deformation and stress redistribution during operation stage.

Understanding the distribution and characteristics of tunnel noise is a prerequisite for noise control and prevention. Tianba Shierpo construction tunnel was established as the prototype, the noise of key regions was measured, the law of noise in key areas was explored. Using discrete wavelet transform, the energy distribution characteristics of the main noise sources in the tunnel were defined, noise hazards are classified. The results show that the frequency spectrum distribution of fan and shotcrete noise is similar, and shotcrete noise is discrete noise. The root cause of the difficulty of noise control in tunnel is the different intrinsic characteristics of different noises. The pattern of energy distribution for fan and shotcrete noise exhibits an initial increase, followed by a decrease, and then another increase, whereas the drilling noise displays a decline initially followed by an increase. It can be graded according to the damage to human body as, drilling > shotcrete > fan.