Support matrix machine is an advanced matrix learning model that can fully utilize the intrinsic structural information in matrix data. However, it is susceptible to noise and outliers, and lacks generalization ability in imbalanced data. To this end, a robust cost-sensitive support matrix machine (RCSSMM) model is proposed and applied to intelligent diagnosis of wind turbine gearbox faults. RCSSMM improves the robustness to noise and outliers by evaluating the prior distribution of the matrix input with assembled matrix distance, and assigning different sample weights to different samples. Additionally, RCSSMM introduces the cost-sensitive loss function that assigns different penalty factors to different categories of matrix data. The optimal values of the penalty factors are adaptively determined with the Harris hawk optimization algorithm to focus on minority class samples and improve the diagnostic performance on imbalanced data. The proposed method is validated using simulated experimental data and real measured data of wind turbine gearboxes. The experimental results demonstrate that the RCSSMM model exhibits more outstanding fault diagnosis performance even under the presence of noise, outliers, and imbalanced data.

Tuned Liquid Damper (TLD) is a simple and effective passive vibration control device. By adding thickening agents to the TLD, the effect of liquid viscosity on the damping ratio and frequency of the TLD system is studied. Firstly, the relationship between thickener concentration and liquid viscosity is measured by a viscometer. Then, rectangular, circular, circular and U-shaped TLD tanks are designed and tested on a unidirectional harmonic excitation vibration table.The influence of parameters on the performance of the TLD is analyzed, such as thickener concentration, water depth ratio of the tank, external excitation frequency, relative excitation amplitude and placement time. Finally, CFD numerical simulation of TLD system is carried out to study the influence of tank size. The results show that increasing the concentration of thickener can effectively improve the damping ratio of TLD, and has little effect on the frequency of TLD. Water depth ratio has little effect on frequency and damping ratio of TLD. The external excitation amplitude and frequency have little effect on the frequency and damping ratio of the TLD system, but can significantly change the liquid surface wave height. The TLD placement time of the thickener liquid increased, resulting in a decrease in the viscosity of the liquid in the TLD, resulting in a decrease in the damping ratio of the TLD, partial volatilization of the liquid in the TLD, and a decrease in the water depth ratio, resulting in a phenomenon of TLD frequency mismatch. As the size of the tank decreases, the TLD damping ratio gradually increases, and the damping ratio remains basically stable when the size is larger.

The seismic performance tests of reinforced concrete (RC) shear walls and BFRP bars reinforced concrete (BFRP-RC) shear walls with different horizontal reinforcement ratios (0.25% and 0.50%) were carried out to explore the similarities and differences in seismic performance between RC and BFRP-RC shear walls. And the horizontal reinforcement ratio was expanded to 0% and 1.00% in meso-scale numerical simulation. The influence of reinforced materials type on the seismic performance of shear walls was discussed, and the shear capacity, deformation capacity, energy dissipation capacity, stiffness and recovery performance of RC and BFRP-RC shear walls were compared. The test results show that the shear failure and compressive shear failure occurred respectively in the shear walls with horizontal reinforcement ratios of 0%~0.25% and 0.50%~1.00% under horizontal cyclic load. The horizontal reinforcement ratio has the same effect on the failure mode, shear capacity, deformation capacity and energy dissipation capacity of RC shear wall and BFRP-RC shear wall; that is, increasing the horizontal reinforcement ratio can enhance the seismic performance of shear walls. However, the seismic performance of RC and BFRP-RC shear walls is different. Under the two horizontal reinforcement ratios, the shear capacity of the BFRP-RC shear wall is about 74%~78% of that of the RC shear wall, the deformation capacity is about 47%~84%, and the initial stiffness is about 77%~84%. Because the BFRP bar is always in the elastic deformation stage during loading, the recoverability of the BFRP-RC shear wall is significantly stronger than that of the RC shear wall. When the horizontal reinforcement ratio is 0.25% and 0.50%, the residual deformation of BFRP-RC shear walls is 62% and 13% of that of RC shear walls, respectively. The recoverability of the BFRP-RC shear wall is more in line with the requirement of recoverable functional aseismic structure in practical engineering.

The present paper proposes an hybrid base isolation system referred to as BRB+NFVD+BIS, consisting of the buckling restrained braces (BRB), nonlinear fluid viscous dampers (NFVD), and base isolation system (BIS) to study both the damping and isolation effects of the hybrid base isolation system on prefabricated high-rise buildings. Defined are the ratios of both BRB yield strength to base isolation yield strength and the total damping index of NFVD to base isolation yield strength, respectively designate as BIR and NIR. Based on the dynamic elastic-plastic seismic response analysis of the corresponding systems, the effects of BIR, NIR, and NFVD parameters on the seismic performance of BRB+NFVD+BIS tall buildings have been revealed, and the ranges of BIR, NIR, and NFVD parameters are suggested. Results demonstrate that with respect to the non-isolated prefabricated high-rise buildings, the BRB+NFVD+BIS system can significantly enhance the seismic performance of beam-column connections, reduce both the inter-story drift ratios and floor accelerations of the superstructure. Compared with the base-isolated prefabricated high-rise structures, the BRB+NFVD+BIS system substantially reduces base isolation layer displacement while maintaining almost the same seismic performance to each other in terms of the beam-column connections, inter-story drift ratios, and floor accelerations Therefore, the BRB+NFVD+BIS system processes better displacement control ability of isolation layer. Simultaneously, the results show that the BRB+NFVD+BIS system has better robustness of both the seismic mitigation and isolation.

There exists the nonlinear failure correlation among the multiple monitoring points of bridge components. Considering the influence of this factor on the reliability indices of the bridge, this paper adopts the Bayesian optimized long short-term memory (BO-LSTM) network model in machine learning to dynamically predict the monitoring data of the bridge, and establishes a three-dimensional Gaussian Copula model based on Copula theory to calculate the time-varying reliability indices and failure probability of the bridge construction. The rationality of the model and method is verified by applying the monitoring data of Fumin Bridge in Tianjin.

This study aims to investigate the life-cycle seismic performance degradation behaviour of reinforced concrete (RC) girder bridges under chloride-induced corrosion. Based on the Duracrete model and existing research results, the time-dependent deterioration models for the mechanical properties of longitudinal reinforcement, transverse reinforcement, cover concrete, and core concrete are determined. A three-span RC continuous girder bridge is taken as an example, and its nonlinear analysis models corresponding to different characteristic time points are established by the OpenSees platform. Four analysis cases are investigated to study the effects of chloride-induced corrosion on the girder bridge's seismic capacity and seismic demand. Among these cases, one involves the omission of considering the deterioration of ultimate tensile strain of reinforcing steel, while the remaining three consider this deterioration using three diverse degradation models. The results show that: in the presence of chloride-induced corrosion, the degradation of the ultimate tensile strain of reinforcing steel manifests markedly more severe than the deterioration observed in its yield strength; the girder bridge suffers a more significant decrease in ultimate curvature, a greater increase in curvature demand, and a lower curvature demand-to-capacity ratio of pier when considering the deterioration of ultimate tensile strain of reinforcing steel; disregarding the degradation of the ultimate tensile strain of reinforcing steel would render the life-cycle seismic performance evaluation results of girder bridge structures unreliable and unsafe; additionally, the applicability of these three deterioration models varies, and there are significant differences in the degree of degradation of curvature demand-to-capacity ratio among these models. Therefore, the choice among these three models should be grounded in the research application scenario. As a result, it is necessary to consider the deterioration characteristics of the ultimate tensile strain of reinforcing steel in the time-dependent seismic performance evaluation of RC girder bridges.

The tire deformation traveling wave characteristics vary at different vibration frequencies and speeds，and are influenced by the coupling of landing gear structural vibration，resulting in continuous changes in tire lateral deformation modes. The tire models for previous traditional landing gear shimmy analysis are quasi-static restoring force model，which assumes that the functional relationship between tire lateral force，self-aligning moment and lateral deformation is fixed. Based on the tire stretched string theory，a time delay tire model coupled with landing gear shimmy is established to analyze the dynamic changes in tire lateral deformation modes，and the real-time lateral deformation distribution of the tire and the nonlinear dynamic behavior of the landing gear during shimmy are obtained. The time delay is introduced to calculate the tire deformation traveling wave，and the stability of the landing gear tire coupled shimmy model is studied by the bisection method. The testing methods for the parameters of time delay tire shimmy model is provided. Using the methods presented in this paper，the landing gear shimmy analysis and experimental test of a certain aircraft are conducted. The calculated vibration frequency and shimmy stability region are consistent with the analysis results of traditional tire models and test results.

Helicopter cabin noise negatively impacts cabin environmental comfort and safety. In this paper，non-dominated sorting genetic algorithm Ⅱ is used to solve the electro-acoustic device placement problem of the active control system applied in the cabin. An appropriate spatial discretization of the sound field of the confined space is preformed，and then the electro-acoustic devices placement optimization problem，quantity optimization problem，and secondary source sound intensity optimization problem are further transformed into a combinatorial optimization problem. Taking the minimun value of the sum of the squares of the acoustic pressures at the measurement points as the control objective，the multi-objective optimization algorithm combined with the active control algorithm is adopted to solve the optimal placement of electroacoustic devices in the system. Considering the noise control system’s complexity and feasibility and limited space inside the cabin，the control system’s secondary sound sources and error sensors are selected as a 4-channel configuration. The optimization program is repeated several times independently，and the most frequently occurring positions of electroacoustic devices are counted，based on which computer simulation and experiment are conducted，and the results show that the optimization results can make the noise reduction at the height of the head of the cabin personnel in the sitting position reach up to 24.9 dB，and the global noise reduction reaches 19.4 dB.

Based on the flow simulations through delayed detached-eddy simulation model，the methods of sparsity-promoting dynamic mode decomposition and spectral proper orthogonal decomposition are applied for analysing the mode decomposition on the wake flow. Then combined by the acoustic analogy approach，this study investigates the behaviour of flow and aerodynamic noise generated around tandem seal-vibrissa-shaped cylinder in comparison with the cases of tandem cylinderlike and elliptical bars with the same characteristic length corresponding to the cylinder diameter of 30 mm for a range of Reynolds numbers （Re=6×10⁴~1.2×10⁵）. Results show that the lift fluctuations of the downstream bars are stronger than those of the upstream bars and the downstream bars dominate the aerodynamic noise radiation. The alternative arrangement of the saddle and nodal planes of seal-vibrissa-shaped cylinder introduces three-dimensional flow separations and suppresses the shear layer interactions，improving greatly the flow stability. The structure destroys the regular vortex shedding of Karman vortex street occurring in tandem cylinder wake. The presence of reverse vortex shedding generated by two adjacent saddle surfaces in the flow of tandem seal-vibrissa-shaped cylinder makes the lateral force balanced partially and reduces significantly the lift fluctuations as well as the vortex-induced vibration. The aerodynamic noise generated by the non-constant fluctuating force exerted on the wall surfaces of bars are suppressed effectively. The sound pressure level is reduced at most frequencies. Thereby the tandem seal-vibrissa-shaped cylinder is demonstrated to have a significant noise reduction effect. The experimental measurements verify the accuracy of the aerodynamic noise predictions. The current work would provide a certain scientific research and engineering application value for the aerodynamic noise control on cylinderlike bars.

Since current loudness models are unable to predict loudness under round window stimulation，a loudness model for round window stimulation is proposed in this paper. The loudness model consists of a peripheral auditory model and a data processing back-end. The peripheral auditory model that is able to calculate basilar membrane velocities under free-field acoustic stimulation and round window stimulation and the back-end that transforms basilar membrane velocities into loudness are constructed. The reliability of the peripheral auditory model is verified by comparing the model-predicted results with the experimental data on the outer ear transfer function，middle ear transfer function and stapes velocities under acoustic stimulation，round window stimulation transfer function，frequency selectivity and frequency response of the basilar membrane，and basilar membrane displacement. The reliability of the loudness model is verified by comparing the model-predicted results with the experimental data on equal-loudness contours，bandwidth noise loudness，loudness level of tone with frequency masking，and threshold for complex tones. The results indicate that the loudness model accurately calculates basilar membrane velocities under acoustic stimulation and round window stimulation，and is able to predict the loudness of pure tone，complex tones，and bandwidth noise under acoustic stimulation and round window stimulation.