The underwater manipulator is the preferred configuration equipment for the operational underwater robot, which has the advantages of flexible movement and strong anti-interference, and has been widely concerned by scholars at home and abroad. A large number of scholars have carried out research on the function, structural characteristics, dynamics performance and control methods of the underwater manipulator. Among them, the dynamics analysis is the research core, but there is a lack of systematic summary at present. Therefore, focusing on the research methods and technologies of underwater manipulator dynamics, the functions and design features of the underwater manipulator used in different underwater operations were summarized; then, the dynamics modeling methods for the underwater manipulator were classified and compared, mainly including the dynamics modeling method based on the error compensation and the dynamics modeling method based on the hydrodynamics model; finally, the construction method of inertial and viscous hydrodynamics models were described in detail. By comparing different dynamics modeling methods of the underwater manipulator, the new research direction in the future was pointed out. The research results can provide some reference for the follow-up research of underwater manipulators.

In the rapid prototyping technology, model segmentation processing is the previous step of layered processing, so it is necessary to study the segmentation algorithm. For this reason, the segmentation algorithms in the rapid prototyping technology were summarized in the aspects of human-computer interaction segmentation and automatic segmentation. For the human-computer interaction segmentation algorithm, the description was mainly based on the segmentation according to different forming requirements; for the automatic segmentation algorithm, it mainly introduced the voxelization segmentation algorithm, the feature recognition segmentation algorithm, the multi-objective optimization segmentation algorithm and the clustering segmentation algorithm. The analysis results showed that there were still some problems in the existing segmentation algorithm, such as limited application scope, single application of rapid prototyping technology, lack of considering subsequent splicing links, and most test models were non-engineering parts. Therefore, carrying out related research on the establishment of evaluation criteria for the segmentation algorithm effect, the application of multiple rapid prototyping processes, and the performance analysis after considering model splicing can effectively improve the practical value of segmentation algorithms and lay the foundation for the subsequent layered processing.

In the computer aided design (CAD), the classification and retrieval of three-dimensional models of mechanical parts facilitates designers to reuse design information, so as to shorten the product development cycle and quickly respond to market demands. For the classification and retrieval of three-dimensional models, a classification and retrieval algorithm based on the polar radius surface moment and hidden Markov model (HMM) was proposed. Firstly, the polar radius surface moments of the three-dimensional model were calculated and the feature vector was formed. After sorting and coding, the feature vector was used as the observation sequence of the HMM. Then, a part of manually labeled three-dimensional models were taken as the training samples, and the HMM was trained by the multi-observation sequence B-W (Baum-Welch) algorithm with scale factors. Finally, the trained HMM was used to classify and retrieve three-dimensional models. Experimental results showed that， compared with the two existing classification and retrieval algorithms, the proposed algorithm had higher recognition rate and retrieval efficiency. The characteristics of the algorithm were as following: the calculation of polar radius surface moment was fast and the voxelization of three-dimensional model was not needed; the HMM had fast training and strong classification ability, and it had certain classification ability without a large number of training samples. The research shows that the proposed algorithm could better solve the problem of classification and retrieval of three-dimensional models, which had certain practical value.

As the core component of vibroseis vibrator, the fatigue reliability of the welding part of the baseplate directly affects its service life and the excitation quality of seismic signal. In order to accurately obtain the fatigue reliability of the welded part of the vibrator baseplate, the reliability of the baseplate was analyzed based on the fuzzy comprehensive evaluation method. Firstly, the set of factors influencing the fatigue reliability of the vibrator baseplate was determined, and the weight matrix and the evaluation matrix of influencing factors were established by the scoring and evaluation of several experts, and then the fuzzy comprehensive evaluation matrix of the fatigue reliability of the baseplate was obtained. Then, according to the fuzzy comprehensive evaluation results, the weight coefficients of the fatigue reliability analysis model of the baseplate based on the S-N curve method and the fracture mechanics method were determined, and a fuzzy comprehensive analysis model for the fatigue reliability of the baseplate was established. Taking the structure, load and material parameters of the vibrator and the influence factors of the fatigue crack propagation as random variables, the fatigue reliability optimization design for the baseplate was carried out. The results showed that the comprehensive fatigue reliability of the baseplate after optimization was improved by 1.8%. The research results can provide important theoretical basis for improving the fatigue resistance and service life of vibroseis vibrator baseplates.

Intelligent manufacturing is one of the key development directions vigorously advocated by the country in recent years. With the continuous establishment, development and improvement of intelligent manufacturing pilots, the digital twin technology has been gradually applied. Based on this, taking the intelligent production line of five-axis machining center as an example, the application of the digital twin technology was summarized. Based on the digital twin technology, the digital model of the physical equipment on the intelligent production line of five-axis machining center was constructed at a ratio of 1∶1 and the corresponding physical attributes were configured, so that the digital model had the same action state as the physical equipment. Then, the signal points of the digital model were set up and connected with the external control signals to realize the virtual debugging of the intelligent production line of five-axis machining center before completion, and then the optimal construction scheme was obtained through the continuous iteration. The practice results showed that the virtual simulation of the combination, debugging and production of the intelligent production line could be realized by the digital twin; based on the virtual simulation function of the digital twin, the visual display of physical attribute parameters and design accuracy of products could be realized, so that the product could pass the adaptability verification in the design stage and realize rapid update; after the completion of the intelligent production line, the digital twin could be used to realize the management of equipment on the production line and the real-time monitoring of its working state, so as to better obtain the operation state of each equipment. In conclusion, the digital twin technology runs through the whole life cycle of intelligent production line, which can be popularized and applied in the field of intelligent manufacturing.

The cloud model uses entropy to control the uncertainty of the cloud drop, and uses the membership function in the fuzzy set theory to judge the quality of the cloud drop. Its forward cloud generator can significantly improve the global optimization ability of intelligent optimization algorithms. Therefore, the forward cloud generator in the cloud model was used to improve the artificial fish swarm algorithm (called cloud model artificial fish swarm algorithm), and it was applied to the fuzzy optimization design of mechanical parts. Firstly, the relevant parameters of mechanical parts required by the fuzzy comprehensive evaluation were determined, and the optimal level cut set was determined by the fuzzy comprehensive evaluation; then, the augmented multiplier method and the cloud model artificial fish swarm algorithm were used to solve the fuzzy optimization problem of mechanical parts, and the quasi-discrete method was used to adjust the parameters of mechanical parts to the corresponding size series, and ensured that they could pass the fatigue strength check. The fuzzy optimization design results of the spur gear and the internal combustion engine valve spring showed that: in order to ensure that the constraint conditions could be met, a larger penalty factor should be taken to make the Hessian matrix in the Lagrange function positive definite; the cloud model could significantly improve the optimization ability of the artificial fish swarm algorithm. The proposed fuzzy optimization method can be widely used in the design of mechanical parts, which has certain significance for the engineering design.

In order to alleviate the congestion of multi-lane stereo garage during parking and picking-up peak, realize the safety, energy saving and time reduction of vehicle access, the optimization design of garage and vehicle access strategy were studied. A multi-lane intelligent stereo garage with accompanying bearing plate was designed, which was divided into two parts: electromechanical execution system and computer control system. The electromechanical execution system was mainly composed of human-vehicle separation cabin, horizontal transmission mechanism, lifting mechanism, stackable storage rack, accompanying bearing plate warehouse and safety protection mechanism. The computer control system was mainly composed of the upper computer (computer server) in the control room and the lower computer (PLC， programmable logic controller) on the roadway locomotive. According to the mechanical design method, an optimization design was carried out on the horizontal transmission mechanism, which eliminated the roll moment of the roadway locomotive caused by the stacking action of the horizontal transmission mechanism, and saved the motor action time. In order to improve the satisfaction degree of car pick-up owners’ waiting for the service, the strategy of "pick-up priority + cross queue access" was proposed, which employed the computer control method to realize the automatic control of pick-up and cross queue access. Genetic algorithm (GA) toolbox of MATLAB was used to optimize the empty space search in parking. The research results provide a realization path for the low energy consumption and low time consumption of lane-type stereo garage.

The key to study the vibration excitation law of seismic vibrator is to accurately describe the interaction between vibrator and ground under sweep signal excitation. According to the elastic half space theory, a vibrator-ground coupling vibration model was established, the mathematical description of vibrator-ground coupling dynamic stiffness and dynamic damping was carried out, the model test was carried out, and the influence law of coupling parameters on dynamic stiffness and dynamic damping was studied. The results showed that the coupling parameters affecting dynamic stiffness and dynamic damping included sweep frequency, soil elastic modulus, soil density and plate bottom area; the interaction between the plate and the ground decreased with the increase of sweep frequency; the interaction between the soil with larger elastic modulus and lower density and the plate was relatively stronger, which was more conducive to vibration output, while the effect of increasing the plate bottom area on improving the dynamic stiffness decreased with the increase of sweep frequency. When using seismic vibrator for exploration, the effects of sweep frequency, soil properties and plate bottom area on dynamic stiffness and dynamic damping should be considered to obtain the desired seismic excitation signal.

The vibration characteristics of the transmission system of large rotating machinery directly affect its smooth operation, and the universal coupling is an indispensable key component in the transmission system. Therefore, taking the tripod-ball cage double universal coupling as the research object, and its torsional vibration characteristics were analyzed. Firstly, the torsional vibration model of the tripod-ball cage double universal coupling was established, and its dynamics equation was established based on the Lagrange equation. Then, the Runge-Kutta algorithm was used to numerically solve the established dynamics equation, and the changes of the rotation angle difference and the angular velocity difference between the input shaft and the output shaft of the tripod-ball cage double universal coupling with time were obtained. Finally, the simulation verification was carried out by the MATLAB software to analyze the effects of the included angle between the input shaft and the intermediate shaft and the external excitation on the torsional vibration characteristics of the tripod-ball cage double universal coupling. The results showed that: the theoretical calculation results of the natural frequency of the tripod-ball cage double universal coupling were consistent with the numerical simulation results; the included angle between the input shaft and the intermediate shaft and the external excitation would affect the amplitude of the rotation angle difference and the angular velocity difference between the input shaft and the output shaft of the tripod-ball cage double universal coupling, which could cause the coupling to produce torsional vibration response; with the increase of the included angle between the input shaft and the intermediate shaft and the external excitation, the torsional vibration response of the tripod-ball cage double universal coupling increased. The research results can provide a theoretical basis for the design, manufacture and application of the tripod-ball cage double universal coupling.

Friction and wear are the main factors leading to the failure of the moving pairs, and the surface texturing is an effective means to improve the tribological properties of moving pairs. Based on the traditional circular and triangular surface textures, several novel surface textures were designed, and their two-dimensional and three-dimensional models were established, respectively; then, the CFD (computational fluid dynamics) software Fluent was employed to simulate the novel surface textures, the effects of geometric parameters (height of arc-shaped protrusion and triangle shape) and distribution angle of the surface texture on its surface velocity field and bearing capacity were studied. The results showed that: for the circular surface textures, those with arc-shaped protrusion at the bottom generally had greater bearing capacity, and the bearing capacity reached the maximum when the protrusion height was 0.6 mm; for the triangular surface textures, the bearing capacity of the novel triangle I surface texture was larger; the distribution angle of surface texture had a significant effect on the bearing capacity. When the distribution angle was 75°, the bearing capacity of the surface texture was maximum. The research results can provide a theoretical reference for the design and optimization of the surface texture of moving pairs.

In order to develop an aircraft with certain load capacity and high concealment, a bionic flapping wing aircraft which can fly at ultra-low altitude was designed according to the flight mode of birds. Firstly, the degree of freedom of the transmission mechanism of flapping wing aircraft was calculated, the feasibility of the design scheme was determined in principle, and the dimensions of each component of flapping wing aircraft were determined. Secondly, the three-dimensional model of the aircraft was drawn by the design software Creo. Through the motion simulation, it was concluded that the flutter of the aircraft met the design requirements. Then, the dynamics simulation of the aircraft was carried out by ADAMS (automatic dynamic analysis of mechanical system) software, and the changes of flapping amplitude and angular velocity in the flapping process were obtained. Therefore, the rationality of the three-dimensional model design of the aircraft was verified, and the balance torque at the rotating joint of the aircraft was calculated. Finally, the prototype of flapping wing aircraft was made, and indoor and outdoor flight tests were carried out. The dynamics simulation analysis and flight tests showed that the designed flapping wing aircraft had good flapping stability, relatively strong flight ability and bionic concealment, and the flight effect reached the expectation. The flapping wing aircraft can fly at ultra-low altitude and has a wide application prospect.

At present, the energy conversion efficiency of passive walking aid exoskeleton is relatively low. Based on the gait characteristic of human body and the gravitational potential energy and kinetic energy generated during walking, the power assisted principle of exoskeleton was analyzed, and a lower limb passive exoskeleton was designed using multi-level energy locking mechanism. Dynamic simulation model of exoskeleton was established, and the energy storage performance of exoskeleton was simulated and analyzed with weight and walking speed as variables. Based on EMG (electromyography) signals, the assistance performance of exoskeleton was studied, and the time-domain curves and integral values of EMG signals of lower limb muscles with and without exoskeleton were compared. The results showed that the multi-level energy locking mechanism could improve the energy storage efficiency of the exoskeleton. The exoskeleton had excellent adaptability to different weight and walking speed, while the stored energy was positively correlated with both weight and speed. The exoskeleton has good boosting effect under different walking states.

As a basic component, bearing has a wide variety and is in great demand in industrial production. For the convenience of users, bearings with different materials at both ends must be sorted before packaging, so that the end faces with the same materials orientate the same direction. At present, most domestic bearing manufacturing enterprises still use manual sorting method, which is inefficient. In order to improve the sorting speed of bearing and reduce the cost of manufacturing enterprises, taking the 608-Z type bearing as an example, starting from the design goal of fast and convenient sorting the bearing, a bearing end face detection and sorting device was innovatively designed, which used the PLC (programmable logic controller) as the control core, the color code sensor as the main sensing element, and the cylinder and the electric rotary table as the main actuator. This device distinguished the orientation of the bearing end face through the color recognition, and controlled the turnover and arrangement of the bearing through the cylinder, which realized the detection and sorting of the bearing. The bearing end face detection and sorting device based on color recognition had the characteristics of fast recognition, simple and reliable mechanical structure and stable operation, which could realize the detection of the bearing end face and the sorting of the bearing. The test results show that the designed device can meet the actual bearing production needs, which can provide a reference for encouraging the transformation and upgrading of bearing manufacturing enterprises and promoting the gradual implementation of unmanned chemical factories.

In order to improve the reliability of micropipetting of biochemical analyzer, a pipetting fault diagnosis and pipetting volume detection system based on image segmentation method was proposed. Based on STM32 microcontroller, with the stepping motor of micropipetting system controlled by speed-position double closed-loop PID (proportion-integral-derivative) algorithm, the automatic pipetting control system of biochemical analyzer was designed. The pipetting process image was collected as the sample image，and the pipetting data set was established. The U-Net neural network models before and after pruning were trained, and the calculation, parameters and mean intersection over union (MIOU) of the models were compared. The pipetting area was segmented and processed by the pruned U-Net model to realize the fault diagnosis of the pipetting system. Combined with the geometric characteristics of the pipetting suction head, the pipetting volume model was established to calculate the pipetting volume. The micropipetting experiment based on image segmentation method was carried out to analyze pipetting errors, and the errors were compensated by least square method, sub-pixel corner point detection and limit learning machine. The results showed that after pruning, the calculation, parameters and MIOU of U-Net neural network model were reduced by 47.30%, 93.99% and 0.61% respectively, and the operation efficiency of the model was significantly improved. For the verification points of 10, 50, 100 μL, the pipetting accuracy of the micropipetting system reached 1.72%, 1.36% and 1.39% respectively, meeting the accuracy design requirements of micropipetting system. The pipetting monitoring system based on image segmentation method can effectively judge the pipetting fault and detect the pipetting volume. The research result promotes the development of micropipetting technology.

Spring assembly is the key component of linear compressor. In order to optimize the spring assembly of the linear compressor, reduce the volume and weight of the spring assembly, and improve the work efficiency of the linear compressor, a multiple linear cylindrical arm coil spring with easy processing and high stiffness was designed. ANSYS finite element analysis software was used to calculate the stiffness and conduct modal analysis on the cylindrical arm coil spring. After that, the influence of the main structural parameters, such as the base circle radius of the coil spring center line, the coil spring line diameter, the coil spring axial height on its stiffness and stress was studied. The results showed that: with the increase of base circle radius, the axial stiffness and maximum stress of the cylindrical arm coil spring gradually decreased, and the radial stiffness variation curve was approximately a quadratic function curve with the opening upward; with the increase of line diameter, the axial stiffness and radial stiffness of cylindrical arm coil spring increased rapidly, and the maximum stress increased linearly. When designing the cylindrical arm coil spring, the base circle radius was set 15-19 mm, that was beneficial to improve the performance of the coil spring; if the line diameter was 4-6 mm, the axial height of 0 mm should be preferred; if the line diameter was 6-8 mm, the axial height of 10 mm should be preferred. The research results have important reference value for improving the spring support technology of linear compressor.

The structure and motion mode of the mobile parallel mechanism have a great impact on its moving performance, while the traditional mobile parallel mechanism can not adapt to the complex terrain environment by switching its motion modes. In order to improve the terrain adaptability of the mobile parallel mechanism, a reconfigurable spatial open/closed chain 6R (rotating pair) mobile parallel mechanism was designed with the planar open/closed chain 6R mechanism as the basic unit. By controlling the opening and closing of the single loop 6R mechanism, this reconfigurable mobile parallel mechanism could realize three motion modes. The folding mode was used for carrying, storage and transportation; in the foot-type obstacle crossing mode, the angle of the rotating pair could be adjusted to adapt to the ground walking; in the rolling mode, the rolling on flat, concave-convex or soft ground could be realized by the alternately spreading and supporting of rods. Through the degree of freedom analysis and drive configuration for the reconfigurable spatial open/closed chain 6R mobile parallel mechanism, the kinematic model of the mechanism was established, and its foot-type obstacle crossing performance and the feasibility of rolling mode in a variety of terrain were analyzed. Finally, the correctness of the structural design and kinematic model of the mechanism were verified by the simulation analysis. The results showed that three motion modes of the reconfigurable spatial open/closed chain 6R mobile parallel mechanism were feasible and effective, and its stability and environmental adaptability were excellent. The research results can provide reference for the design of reconfigurable mobile mechanisms.

In view of the large size of the muffler used to reduce the broadband low-frequency noise, and its poor noise acoustic attenuation performance, based on the broadband acoustic attenuation characteristic of the perforated tube and the good acoustic attenuation performance of the membrane structure for low-frequency noise, an acoustic attenuation structure of perforated tube coupled with metamaterial film was designed. The acoustic attenuation mechanism of acoustic attenuation structure was analyzed, and the influence of the structural parameters on acoustic attenuation performance was analyzed by simulation. The results showed that the higher the coupling strength between the acoustic wave in the main pipe and the resonance system of acoustic attenuation structure, the stronger the vibration of the film, the stronger the acoustic wave reflected back to the upstream pipe, and the weaker the acoustic wave transmitted to the downstream pipe; the frequency corresponding to transmission loss peak of acoustic attenuation structure was closely related to the first resonant frequency of metamaterial film. The spiral acoustic attenuation structure was designed, and its performance was simulated and tested. The results showed that the simulation result was in good agreement with the experimental result. The research results provide a useful reference for the control of broadband low-frequency noise.