Taking aircraft brakes as the research object，the problem of aircraft brake induced vibration was studied. Aiming at the influence of disc friction characteristics caused by the complex braking environment during aircraft skidding，a nonlinear dynamic analysis model of brake disc set was established considering the properties of the hydraulic system and disc material，with the model interface reserved for the development of new material technology. From the perspective of practical engineering research value，the influence of the above parameters on the stability of the system is studied by using numerical simulation，bifurcation theory，and other nonlinear analysis methods by considering key factors such as wheel speed，hydraulic system equivalent damping，brake disc surface damping and brake disc friction characteristics as control variables，and the range of structural parameters to keep the system stable is determined. It is convenient to parameterize the structure for rapid analysis and design. Based on the results of the model analysis，part of the causes of brake vibration were revealed，and corresponding optimization measures for vibration reduction were put forward from two aspects of structural parameters and brake control law. Then a comprehensive analysis method for aircraft brake induced flutter was formed，which provided a theoretical reference for the initial stage of aircraft brake design.

At present，many rolling bearing fault diagnosis methods based on convolutional networks have the disadvantages of poor diagnosis effect and poor generalization ability under the influence of noise signals and load variations. Aiming at these problems，an improved convolutional capsule network fault diagnosis method of rolling bearing under variable operating conditions is proposed. This method designs a multi-scale asymmetric convolution module，in which asymmetric convolution layers of different scales to extract features from the input data to maximize the extraction of feature information in the data and reduce the number of parameters effectively. In this module，the channel attention mechanism is introduced to better extract useful channel features and improve the feature extraction ability of the method in this paper. By improving the fully connected layer in the network to the fully connected layer of the capsule，the capsule can avoid the loss of characteristic information in the space in the process of outputting vector feature information. Case Western Reserve University bearing dataset and Southeast University gearbox dataset are used to verify the diagnostic performance of the proposed method and compare with other deep learning methods. The experimental results show that the proposed method has a better generalization and performance.

The band gap characteristics of periodic structures provide a new idea for the field of seismic isolation in civil engineering，among which the one-dimensional periodic foundation structure has garnered significant attention due to its simple structure and economical applicability. In this paper，by studying the vibration characteristics of the one-dimensional periodic base structure，an approximate analytical solution for calculating the one-dimensional rubber-concrete periodic base band gap is derived，and on this basis，a one-dimensional rubber-concrete periodic foundation optimization design method based on the resonance zone of the superstructure is proposed. Numerical examples in the frequency domain and time domain show that the periodic foundation designed by this optimization method can ensure a good damping effect of its superstructure in a wide and continuous frequency range.

The shortcomings of fault diagnosis methods based on deep convolutional neural networks is that，tensor data is easily destroyed when reducing the dimension of high-order input tensors by pooling layers，which results in a loss of data information，and the relatively complex network structure. Therefore，a Deep TensorProjection Networks method is constructed via replacing the pooling layer in the traditional CNN by a TensorProjection Layer. The TensorProjection Layer reduces the dimensionality of input high-order tensor data without causing damage to the data，thus avoiding the impact of the loss of feature information，and greatly improving the recognition accuracy of the model. The dimensionality of the TensorProjection Layer used for dimensionality reduction is variable，thus simplifying the networks structures. Based on this，combined with the respective advantages of high-order spectrum and deep TensorProjection networks，a mechanical fault diagnosis method based on deep TensorProjection networks is proposed. In the proposed method，the feature of fault signal is extracted by high-order tensor spectrum，which is input into the constructed model for reducing high-order tensor dimensionality and identifying faults. The proposed method is applied to diagnose gearbox faults. Experimental results show that the proposed method can better retain the original fault information and effectively recognize the different types of faults. And the accuracy is better than traditional deep convolutional neural network fault diagnosis methods.

There is a lack of detection means for the isolation performance of the buildings built by passive control technology，so it is of great significance to test the dynamic characteristics of the isolated structures on the spot. A 4-story base isolation kindergarten was tested in the field，and the test device，method and results were displayed. The results were compared with the seismic structure model under the same conditions，and the dynamic response law and damping effect of the actual isolation structure were explored. The building was pushed away with hydraulic jack to produce 98 mm （corresponding to LNR500 shear strain 102%） horizontal initial displacement of the isolation layer，and concrete jacking rod was installed to support the building； The concrete rod was blasted with explosives and unloaded instantly to make the building vibrate freely； The dynamic response and other parameters were tested and analyzed. The results show that under the condition of horizontal initial displacement，the first-order natural vibration period of the isolated structure is significantly longer than that of the seismic structure，and the damping ratio increases； The hysteretic curve of the isolation layer is full； The dynamic response control effect of each floor is obvious，but the acceleration of the roof floor is slightly amplified compared with that of the bottom floor； After unloading，the isolation layer instantly resets，which shows that the isolation layer has rapid reset performance.

To address the threat of harmful vibrations of semisubmersible floating offshore wind turbine （FOWT） in complex deep-sea environments to the safety and durability，a design of distributed tuned mass dampers （TMDs） is proposed to control the platform pitch motion under the randomly combined wind and wave excitations，in combination with the geometric structure of the 5MW prototype of NREL in the United States. The distributed TMDs are installed inside the platform to form an equilateral triangle arrangement. To better describe the performance of the distributed TMDs on the semisubmersible FOWT，a 9-degree-of-freedom multi-body dynamics model is proposed and established for the coupled semisubmersible FOWT-TMDs system based on Lagrange's equation and the modal superposition method. Based on the H_∞ algorithm，whose optimization objective is the peak value of the frequency response function of platform pitch motion，the parameters of the distributed TMDs are optimally designed，where the coupling relationship between multiple TMDs is considered. The numerical simulation of the coupled FOWT-TMDs system under the combined wind and wave excitations is carried out to analyze the performance of the distributed TMDs on the platform pitch response of the wind turbine. The results show that the distributed TMDs with optimal design has good damping performance on the platform pitch motion of the semisubmersible FOWT. Under random wind and wave loads in three different working conditions，the peak and standard deviation vibration reduction rates of the power spectral density curve near the natural frequency of platform pitch can reach more than 39% and 52%，respectively. The research method and results can provide reference for dynamic analysis and vibration control design of large semisubmersible FOWT.

Most of the nuclear power plants are built around the coastal or along the rivers. Below the underground water level，distribution of the water in the soil pore has a great influence on the seismic response of soil，which affects the response of the nuclear power plants. To analyze the effect of the underground water level on the seismic response of nuclear power plant，a saturated porous medium model considering the interaction of the saturated soil and structure is used in this paper. Firstly，the free field of the horizontal layered site of dry soil-saturated soil is obtained by the transfer matrix method，and the wave input of soil-structure interaction analysis is realized combined with the transmission boundary； Then，the partitioned parallel calculation method of soil-structure interaction is used to analyze the saturated soil-structure interaction. The soil with groundwater level is described by the generalized saturated porous medium model which is simulated by the lumped-mass explicit finite element combined with the transmission boundary using the self-programmed FORTRAN code，and the structure is analyzed by ANSYS using implicit finite element. Taking a nuclear power plant as an example，the dynamic response of soil-base-nuclear power plant system is analyzed in five sites with different groundwater levels of -10 m，-20 m，-30 m and -40 m，as well as pure saturated soil. The results show that the groundwater level has a great influence on the response of foundation and structure. For the calculation example in this paper，the results show that the groundwater level has little effect on the displacement of the foundation and structure，but has a great effect on the acceleration of the foundation and structure.

To reveal the influence of biaxial coupling effect on the multi-dimensional seismic performance of flanged reinforced concrete （RC） shear walls with different section forms，three T-shaped and two L-shaped RC shear walls were tested under low cyclic loading along their principal axes. The failure modes，hysteretic characteristics，bearing capacity，ductility，ultimate drift ratio，energy dissipation capacity and reinforcement strain of RC shear walls with flange under uniaxial and biaxial lateral loading were compared and analyzed. Test results show that failures of T-shaped walls and L-shaped walls exhibit obvious asymmetry，and the damage is concentrated at the free end of wall penal. Biaxial loading aggravates the cracking and damage degree of RC shear wall with flange，and is likely to cause local damage concentration. Compared with the RC shear wall with flange under uniaxial loading，the biaxially loaded specimens have smaller bearing capacity and deformation capacity in all directions，larger proportion of flexural deformation in the plastic hinge area of web segment，faster energy consumption，poorer energy dissipation capacity in a single direction，larger strains of vertical reinforcement in web and flange，and more obvious shear lag effect of flange. Biaxial coupling effect has more pronounced influence on the damage evolution of L-shaped walls than that of T-shaped walls，resulting in greater reduction of seismic performance index of L-shaped walls under biaxial loading than that of T-shaped walls. Considering the biaxial seismic actions，the limit value of inter-story drift ratio of RC shear walls in China's seismic design code is still relatively safe，but the safety redundancy is reduced.

In order to give full play to the seismic potential of the movable support pier and improve the overall longitudinal synergistic effect of the continuous girder bridge，based on the principle of functional separation and synergistic force，a new type of mass rotation wrap rope device is proposed based on the mechanism of wrap rope. Taking a typical three-span continuous girder bridge as an example，the shaking table test is carried out by inputting actual seismic waves with different seismic spectrum characteristics and intensity as excitation，the seismic response with equal pier height model and unequal pier height model are analyzed to explore the synergistic force and shock absorption effect of the device on the continuous girder bridge. Through the test results of the response of the key positions of the structure such as the acceleration response，displacement response and strain response，it can be seen that the effect of the device on the movable bearing pier participating in the overall longitudinal synergistic force of the continuous girder bridge is more obvious，and with the increase of the ground motion input intensity，the synergistic effect of the device becomes more and more prominent，the design intention of the device is realized. At the same time，the effect of the device is related to factors such as the number of wrap rope turns of the device itself，the pier height of the movable bearing pier，etc. The design needs to determine the reasonable design parameters of the device according to different factors such as the pier height to achieve the best effect of the device.

Stay cables of cross-sea cable-stayed bridges will generate self-excited vibrations in non-extreme wind environments. Based on technologies of modern monitoring and data analysis，the digital features of the self-excited vibration of the stay cable can be captured immediately，and the real-time performance of dynamics about stay cables can be reflected accordingly. According to features of commonality in the long-term monitoring data of vibrating acceleration of stay cables from a main channel cable-stayed bridge of a cross-sea bridge，an automatic extraction method for the non-stationary sections of wind-induced self-excited vibration of the stay cable is proposed based on the vibrating acceleration time series signal’s Gaussian mixture model of the upper envelope and the power spectrum of the frequency domain. The strategies，that identify the dominant frequency based on the non-stationary time series of self-excited vibration，then identify the damping ratio using the data of the last descent section after band-pass filtering，are proposed. By the proposed strategies，the interference on the damping ratio identification from energy brought by the natural excitation of the ambient wind in the ascent sections of the vibration amplitude is excluded. Based on the identified results of dominant frequency-damping ratio in the non-stationary sections of the self-excited vibration，data clusters of the modal frequency-damping ratio corresponding to each order of the vibrating mode are obtained by clustering the frequency values. According to the discrete and skewed characteristics of the damping ratio data，the statistical law of using the eμ of log-normal distribution model and the quantile value of its cumulative distribution function to describe damping ratio of stay cables is proposed. The frequency centroid of each cluster，as well as the probability characteristic parameter of damping ratio are used as indicators to represent the current state of the dynamic performance of stay cables. The main conclusions for the background engineering include： the signal of vibrating acceleration of stay cables on the bridge has strong noise and large interference，they must be eliminated in the analysis of dynamic characteristic； the amplitude maximum of the acceleration of the self-excited vibration of the bridge’s stay cable has exceeded 3000 mm/s²，the vibration amplitude is large； the average level of the damping ratio of the bridge’s stay cable is about 0.03%，which is lower than the recommended value of the design code.