Ginsenoside Rg5 is a conversion product of the protopanaxadiol saponin (PPD), exhibiting good pharmacological activities such as anticancer, anti-inflammatory, memory improvement, and anti diabetes. Due to its poor stability, ginsenoside Rg5 is commonly prepared employing mild organic acids as catalysts, a practice that may result in diminished reaction rate and yield. This study aims to optimize the process conditions for the preparation of ginsenoside Rg5 catalyzed by hydrochloric acid assisted by high viscosity chitosan, with the objective of enhancing the conversion rate of ginsenoside Rg5 and minimizing its decomposition. The surface response method was employed for the optimization process. The preliminary exploration of the mechanism of action of chitosan was conducted. The experimental findings indicate that the acid concentration of 0.17 mol/L, high viscosity chitosan dosage of 20 mg, reaction temperature of 65 ^oC, and reaction time of 100 min constitute the optimal process for the preparation of ginsenoside Rg5. It was observed that under the optimal conditions, the highest yield of Rg5 achieved 58.92%. The investigation into the adsorption mechanism of high viscosity chitosan reveals that the initial adsorption process involves the binding of hydrogen protons by chitosan, followed by the subsequent adsorption of PPD saponins facilitated by the hydrogen protons. Furthermore, it is observed that high viscosity chitosan exhibits a greater affinity for strongly polar saponins compared to weakly polar saponins. The high viscosity chitosan facilitates the ongoing adsorption of PPD saponins, which possess greater polarity, onto the surface of chitosan. This process leads to the formation of ginsenoside Rg5, which exhibits lower polarity. Subsequently, the generated ginsenoside Rg5 is dissolved in a solution with reduced acid concentration, aiding in the stabilization of Rg5. This sequence of events involving adsorption, conversion, and desorption forms a cycle that enhances the yield and rate of ginsenoside Rg5 preparation.

The extraction and preparation of total saponins from Shenmei Wuwei granules were studied, and the activity in vitro of the extract was explored.Single factor and orthogonal experiment were applied to optimize the extraction technology of total saponins from Shenmei Wuwei materials and the preparation technology of granules.The content of total saponins in extract was determined by ultraviolet spectrophotometry, and the activity in vitro of the total saponins extracted from Shenmei Wuwei granules was discovered by α-glucosidase inhibition rate and DPPH free radical scavenging. The optimum extraction technology was obtained as follows: extraction solvent, temperature, extraction time, material-to-liquid ratio and extract times were 70% ethanol solution, 80 ℃, 1:10 (g/mL), 2.5 h and 2 times respectively which offered the extraction rate of total saponins from Shenmei Wuwei materials to 7.36 mg/g. The preparation process of Shenmei Wuwei granules was as follows: the mass ratio of extract powder to diluent was 1:1 (g:g), the mass ratio of lactose to dextrin in diluent was 2:1 (g:g), sucralose was used as the flavor correcting agent, and the wetting agent was 85%. ethanol solution. Activity test results showed that the total saponins of Shenmei Wuwei had good hypoglycemic and antioxidant activities in vitro. Shenmei Wuwei granules provides basis for its further research and development in health food field.

In order to control the top pressure of ammonia removal tower in the morpholine process, a control system for selecting the top pressure was designed based on Siemens S7-300 PLC. The top pressure is controlled by condensate flow rate in normal operation. When too much non-condensing gas accumulates in the reflux tank，leading to the pressure regulating function of the condensate fails, the non-condensing gas is discharged through the venting control loop, and then the pressure regulating ability of the condensate is restored. The off-line simulation test shows that after putting in automatic control, the control program can satisfy the predetermined control requirements and switch the controller according to the limited parameter conditions. It has a good reference significance for the actual engineering transformation on the spot.

In order to effectively prevent confined space accidents in the chemical industry and reduce casualties, based on the statistical data of confined space accidents in the chemical industry, the rules of accident occurrence are clarified. Apply the 24Model system to explore the causal factors of the "4.15" major fire poisoning accident, and deeply explore the internal correlation between the factors, calculate the total input and output degree of each causal factor, and thus clarify the priority of preventive measures. The researchs result indicate that the number of confined space accidents in China's chemical industry showed a stable trend of change, with the majority of accidents classified as major accidents (accounting for 65.6%), and the main types of accidents were poisoning and suffocation (accounting for 59.1%). Among the factors causing the "4.15" major fire poisoning accident, the total input and output of LMZ refrigerant synergist incorrectly stored in the workshop basement is the highest (C (B1)=0.44), which has a priority level of pre-control. The research results provide data support and strategic reference for the development of accident prevention measures in confined space in chemical industry.

SiC/Si3N4 ceramic composites were prepared by hot pressing sintering, and the effect of temperature and pressure on the mechanical properties and microstructure of the composites were studied. The results show that: (1) With the mass ratio of Si3N4: SiC: CaF2= 6:4:1, the bending strength of the material at 1400℃ and 25MPa reaches the maximum of 66.7MPa, the bulk density is 2.2g/cm3, and the porosity is 29.01%. As the temperature rises, the properties of the material gradually decline. At 1400℃, the main crystalline phase is still β-Si3N4 and contains a small amount of SiC. (2) Under the same mass ratio, sintering temperature of 1400℃ and the sintering pressure of 45MPa, the mechanical properties of the composite can reach 77.4MPa, the bulk density can reach 2.3g/cm3, and the porosity can reach 27.09%. It can be observed from the SEM diagram that the density and mechanical properties of the material gradually increase with the increasing of sintering pressure.

Inspired by the research on the structure of upper limb rehabilitation robot at home and abroad, a new type of upper limb rehabilitation robot is designed according to the structure and movement form of human body. The rehabilitation robot is a hybrid drive upper limb exoskeleton rehabilitation robot, which can realize upper limb rehabilitation under the condition of human sitting posture. First, the initial modeling and design of the upper limb rehabilitation robot are carried out. Secondly, the motion simulation of the preliminary modeling is carried out to observe whether the preliminary modeling design can carry out the corresponding rehabilitation actions. Then, the detailed structural design of the rehabilitation robot is carried out according to the preliminary modeling model, and the reduction gear and motor used in the robot structure are calculated and selected. According to the detailed structure, the rehabilitation robot is simulated to simulate the relevant actions of rehabilitation training, and verify whether the rehabilitation robot meets the rehabilitation requirements. Finally, the finite element analysis of the rehabilitation robot is carried out, and the parts of the rehabilitation robot are optimized.

Curved high-pressure pipeline is widely used in engineering practice. To analyze its transport characteristics, by constructing a dynamic model. The finite element model of multi-angle pipeline with fixed fillet was established by SolidWorks software, ANSYS Workbench platform builds Fluent module and transient structure module for fluid-structure coupling, create data transfer for modal analysis, then the relationship between the impact, stress and strain on the pipe wall at the bend after the high-pressure fluid flows into the pipe is studied. The results show that the influence of different angles on the pipe wall is different, which provides theoretical reference for engineering practice to reasonably avoid the harm of pipeline pressure.

The S-type pipeline is used in the complex terrain during transportation, and different degrees of erosion corrosion will be formed at the elbows of the upstream and downstream, resulting in safety risks. In order to analyze the influencing factors of corrosion and establish the S-shaped pipeline corrosion model, the Fluent module was built on the Workbench platform of ANSYS to study the corrosion models under different influencing factors, create data transmission, and then study the relationship between various factors and erosion rate after the fluid flows into the bend. The results show that the scour corrosion of S-shaped pipeline is affected by many factors, especially the scour corrosion at the downstream elbow. Appropriately reducing particle diameter and mass flow rate, and increasing the ratio of pipe diameter to bend diameter can slow down the corrosion and reduce the corrosion area. The research can provide theoretical basis for safe and stable operation of pipeline corrosion.

A modified YOLOv5s detection algorithm has been proposed to address the issues of slow classification speed, error-proneness, and low flexibility in warehouse goods categorization. The algorithm aims to pre-classify warehouse goods. Firstly, based on the external characteristics of warehouse goods, they are divided into two main categories: packaging boxes and packaging bags, forming a training dataset. Secondly, the backbone network is replaced with MobileNetV3, a smaller-sized model, to accelerate inference. Additionally, an SE attention mechanism module is added to enhance the detection accuracy of the model. Finally, the α_CIoU loss function is incorporated to improve the flexibility of the model. Experimental results demonstrate that the improved algorithm achieves a 2.1% increase in precision (P), a 0.5% increase in mean Average Precision (mAP), and a 10.6% increase in Frames Per Second (FPS) compared to the original algorithm. It enables efficient pre-classification of warehouse goods.

In order to improve the absolute positioning accuracy of industrial robots, a method based on DBO-BP and offline feedforward correction is proposed. This method is suitable for the research on positioning error compensation of industrial robots. By using Latin Hypercube Sampling method to obtain the pose samples of industrial robots, and using BP neural network to establish an error prediction model, the DBO optimization algorithm is applied to improve the local optimal phenomenon, thus improving the convergence and robustness of the model. After offline feedforward compensation processing, the positioning error of industrial robots is reduced, and the absolute positioning accuracy of robots is greatly improved. This method can effectively improve the accuracy and stability of robots, and provides a feasible solution for the precise positioning problem of industrial robots.

In view of the problem that the tower bottoms temperature in the morpholine distillation process is easily disturbed by pressure fluctuations, a study on a cascaded PID control scheme was conducted. The controlled object transfer function was obtained through system identification of on-site production data. A simulation model was constructed using Simulink to validate the feasibility of the cascaded PID control scheme. Subsequently, a cascaded control system for kettle temperature was designed based on the S7-300 PLC. This control scheme is characterized by simplicity, ease of debugging, and straightforward implementation within the PLC system. It holds significant potential for widespread application in view of its favorable attributes.

 In order to improve the output performance of the piezoelectric pump, a four-chamber hybrid piezoelectric pump design is proposed by combining the advantages of piezoelectric pumps with series-parallel structure. Driven by sinusoidal alternating current (AC) in the range of 110V, 40Hz~400Hz, the fabricated test prototype is subjected to the conveying performance tests of gaseous medium and liquid medium. The results show that: when pumping air, the piezoelectric vibrator working mode has negligible effect on the output performance, the maximum output flow rate can reach 3036 ml. When pumping liquid water, the asynchronous drive is used, the output performance is better, and the output flow rate reaches a maximum of 1067 ml when the driving frequency is 140 Hz. The test results provide some reference for the practical application of multi-chamber mixed-connection piezoelectric pumps.

In the production process of traditional brake discs, manual labor is almost always used for loading and unloading the machine tools. This operation method has the problems of low handling efficiency, high production cost, and cannot meet the requirements of efficient production. A loading and unloading system for brake disc workpieces based on machine vision is proposed. It combines vision with industrial robots and builds a system that can accurately identify the pose of the workpiece is obtained, and the pose result is converted into the coordinate system of the robot, thereby guiding the robot to accurately grasp the workpiece. Experiments have proven that the system has autonomous flexibility and can successfully complete the loading and unloading of brake discs.

The internal combustion engine of rail transit is mainly responsible for shunting operations in the depot, online maintenance, online emergency rescue, line and contact network detection, and other work. Internal combustion engines are indispensable equipment to ensure the safe operation of rail transit. When operating in subway tunnels, internal combustion engines produce a large amount of exhaust gas that cannot be discharged from the tunnel, causing great pollution to the air inside the tunnel, which is very unfavorable for maintaining workers' online operations. Based on the above reasons, a tail gas water washing tower has been designed for use on diesel locomotives, and good results have been achieved in actual operation.

With the support of LiteOS real-time operating system, the overall functionality of embedded system needs to be achieved through the collaborative operation of multiple tasks and interrupt service program ISR. The serial communication port, as a shared resource, will be accessed concurrently by two or more tasks or interrupt service programs (ISR). To ensure the reliability and integrity of serial data communication, a universal frame structure design for serial data communication is proposed. Through in-depth analysis of the problems in serial data communication under the LiteOS system, an efficient and stable method for serial data transmission and reception under the LiteOS system is proposed.