Abstract:The sparse solution to linear equations has been widely used in image reconstruction， signal processing， and machine learning. By introducing l1 norm regularization， it can be transformed into solving a constrained optimization problem. Based on a novel probability criterion for selecting the working rows from the coefficient matrix， a sparse greedy randomized Kaczmarz method was proposed， and the convergence analysis of the novel method with and without noise interference were conducted， which showed that the convergence factor of the novel method was smaller than that of the randomized sparse Kaczmarz method. The numerical experiments verified the effectiveness of the proposed method.

Abstract:An alternative greedy Gauss-Seidel method for solving the large linear least squares problem is proposed based on a novel strategy for choosing the working columns of the coefficient matrix A， and the convergence of the new method is analyzed. The numerical experiments show that for the same accuracy， the method proposed in this paper outperforms the greedy randomized coordinate descent method proposed recently in term of the computing time.

Abstract:Linear discrete ill-posed problems arise from many scientific computation and engineering application areas. Considering the solution to large scale ill-posed problems with box constraints， A novel class of randomized internal and external iterative methods are proposed based on the active set strategy， which contain a two-level iteration that， for the external iteration， updates the active set and free variable set， and orthogonally project the iterate onto the feasible boundary， while for the internal iteration， solves the unconstrained linear system with the Krylov subspace methods. The proposed novel active set algorithm is to utilize the efficient randomized method for the internal iteration and the step size is chosen by the Armijo criterion， so that the objective function value can be monotonically decreased with the increase of the number of iterations. Numerical experiments on the image restoration show the efficiency of the proposed algorithm. Under the condition of the discrepancy principle， the computational complexity， iteration steps， and CPU time of the novel active set randomized iteration methods are less than those of previous methods.

Abstract:Based on the sequential quadratic programming algorithm， a class of active set methods for solving real symmetric complementary eigenvalue problems is constructed in this paper. By designing a special strategy with the active set index selection， the iterative sequence generated by the active set method has the characteristics of monotonous decline， and the convergence of the method is theoretically proved. The numerical experimental results show that the method is effective and superior to built-in algorithm of MATLAB in complementarity and iteration time.

Abstract:To describe the viscoelastic characteristics of saturated soft soils， the fractional Merchant model is introduced， and the stress-strain relationship in the transformed domain is derived through integral transforms. Based on the elastic-viscoelastic correspondence principle， the solution of the fractional transversely isotropic viscoelastic saturated soft soils is obtained， which is the kernel function of the boundary element solution for the soils. Based on the stiffness matrix of a 2-noded pile element subjected to axial loading， the finite element solution for the pile is constructed. The boundary element solution for the soils is coupled with the finite element solution for the pile to solve the interaction between the soils and piles. Several examples are designed to verify the presented theory and to analyze the influence of the fractional order on the pile-soil interaction.

Abstract:To assess creep-induced delayed compaction localization in rate-dependent porous rocks， a theoretical criterion for compaction localization is proposed based on elastic-viscoplasticity and the theory of controllability. By establishing systems of local ordinary differential equations， the elastic viscoplastic constitutive operator is defined under the stationary external perturbation condition to identify the unstable acceleration deformation inside the spontaneously propagating strain localized bands. By detecting the loss of controllability in the systems， an instability index for compaction localization in rate-dependent porous rocks is derived. The proposed index is verified by simulating the quasi-instantaneous and delayed compaction localization in a porous rock based on the viscoplastic constitutive model. Furthermore， the applicability of the proposed index in analyzing delayed compaction localization in boundary value problems is validated by using finite element simulation of plane strain compression tests.

Abstract:Concrete-filled circular steel tubular （CFST） pier has been widely applied to urban viaducts in China due to its superior features of less road occupation， higher bearing capacity， and better energy-dissipation ability， yet the practical seismic design method is not available in earthquake prone areas. Considering the research status and deficiency of the concrete filling rate of circular CFST piers， the minimum concrete filling rates are proposed in this study based on different seismic design strategies （seismic isolation design or ductile seismic design） specified in the Bridge Seismic Design Code. In addition， corresponding practical formulas for the concrete filling rate are proposed and verified by theoretical study and numerical simulation， which provides specific design guidance for the seismic design of the circular CFST piers of urban viaducts in high earthquake-intensity areas.

Abstract:The single pile foundation with restraint plates （single hole， four holes） installed inside the traditional steel pipe piles is a new type of foundation for offshore wind turbines. In order to study the lateral bearing mechanism of the innovative pile foundation with restraint plates， a lateral loading system is specially designed in the centrifuge. A series of centrifuge tests are conducted with the traditional piles and the piles with restraint plates under ultimate conditions. The finite element model is developed in ABAQUS and validated against the test results to study the failure modes of the piles. The evolution of the rotation center on the piles are revealed in the numerical analysis. The results show that the innovative piles with restraint plates demonstrate a better stability in terms of stiffness and strength. The lateral capacities are significantly enhanced by adding the restriction plates.

Abstract:In this paper， a systematic study on the failure mode of framed PVB laminated glass， which is commonly used in practical engineering， is conducted based on dynamic nonlinear finite element analysis and field blast test. First， a fine finite element model for simulating framed PVB laminated glass subjected to blast loading is developed using LS-DYNA. Based on numerical analysis， three different failure modes， i.e. global failure， local （punching） failure， and mixed failure， are identified for framed PVB laminated glass in different explosions. The failure mode is found to be dependent on the residual energy ratio of the laminated glass at PVB rupture. Furthermore， the identified failure modes are verified by a series of field blast tests. The results of this study can provide reference for practical engineering blast resistant design of framed PVB laminated glass curtain wall.

Abstract:The finite element model updating technology is widely used in the mechanical and other fields. Because of the influence of many factors， there are many uncertain errors between the actual structure （such as tower crane） and the finite element model， resulting in the distortion of the finite element analysis results. Therefore， it is of great significance to study the finite element model updating with uncertain parameters. In this paper， the finite element model of a tower crane is updated considering the uncertainty of parameters. In order to improve the efficiency of model updating， the response surface model is introduced to replace the finite element model of tower crane. Considering the fact that the RBF neural network has the advantage of high precision fitting for complex problems， an improved interval inverse response surface method is proposed to update the tower crane with uncertain parameters. The feasibility of this method is proved by an example of spring-mass calculation， and the actual tower crane structure is modified. This method improves the deficiency of the interval inverse response surface method， whose result has a good calculation accuracy and efficiency.

Abstract:Three models were developed to predict the air-conditioning hourly cooling load of a subway station from the aspects of optimization algorithm ensemble back propagation neural network （BPNN） and BPNN with data clustering pre-processing. The results show that the influence of the same physical parameters on the air-conditioning load of the subway station reflects a certain dynamic change characteristic over time. Quantitative analysis of these features based on historical data is of great benefit to precisely selecting the model input parameters and improving model prediction accuracy. In the three given models， the predicted mean absolute percentage error （MAPE） of particle swarm optimization （PSO）-BPNN and the fruit fly optimization algorithm （FOA）-BPNN decreases by 25.87% and 40.08% respectively compared with that of BPNN， while the MAPE of Kmeans-BPNN is reduced by 61.12% and 51.90% respectively compared with that of PSO-BPNN and FOA-BPNN， which means that the performance of optimization algorithm ensemble models is better than that of pure BPNN on even ground. Moreover， BPNN with data clustering is better than optimization algorithm ensemble BPNNs after distinguishing the characteristics of real load changes.

Abstract:Aimed at the problem of insufficient adhesion between the wheel and the rail in the further development of high-speed train， an electromagnetic supercharging device was proposed based on the principle of electromagnetic action. An electromagnetic coil was used to generate the suction between the wheel and the rail to increase the adhesion of train. First， the basic structure model of electromagnetic supercharging device is established and the electromagnetic suction was calculated. Then， the effects of train speed， excitation current， winding coil circle number distribution and shape of coil， and coil height from rail surface on electromagnetic suction were simulated and analysed. The electromagnetic testing apparatus was developed to verify the exactness of the model and obtain the variation law of electromagnetic suction with speed， current， coil height from rail surface， and coil shape. The results show that the electromagnetic supercharging device can meet the requirements of increasing adhesion and degaussing wheels at different speed stages and can add reverse excitation current to the wheel for demagnetization when necessary. The research shows that the stable control of the electromagnetic suction of each wheel of the bogie can be achieved by reasonably adjusting the excitation current， which can effectively improve the stability and safety of train operation.

Abstract:Aimed at the hydrogen demand during the actual operation of the onboard methanol reforming hydrogen production fuel cell vehicle， a control-oriented hydrogen supply strategy is designed in this paper. First， based on MATLAB/Simulink， a methanol reforming hydrogen production fuel cell system and a vehicle simulation model are built. Combining the dynamic power requirement for the fuel cell in the actual working condition， with the goal of the lowest methanol consumption， a time series index prediction algorithm is proposed to predict the hydrogen consumption rate of the fuel cell in advance， and then adjust the methanol supply of the reforming system in real time. The simulation result of the C-WTVC cycle shows that this strategy can reduce the comprehensive equivalent alcohol consumption by 1.47%. In addition， taking into account the fact that frequent changes in vehicle cycle will reduce the efficiency of methanol reforming， a rule-based hydrogen supply management strategy is further designed to maintain the methanol reformer operating in a high-efficiency zone， and the comprehensive equivalent alcohol consumption under C-WTVC cycle is reduced by 3.82%.

Abstract:The effectiveness and necessity of adding International Reference Ionosphere （IRI） model constraints and different data selection strategies are determined in the global ionospheric modeling with the carrier-to-code leveling method. The experimental analyses show that by using the original dual-frequency P₁ and P₂ observations with IRI model constraints， the overall effect of ionospheric modeling is significantly improved at some epochs. Specifically， in those areas where observations are insufficient， the positioning errors are significantly reduced by 0.260 m， 0.146 m， and 0.103 m in the North， East， and Up directions respectively， which verifies the effectiveness of the method. After replacing the missing P₁ observations with C₁ observations as low-accuracy data， the overall quality of ionospheric modeling is further improved， although the model accuracy is slightly lost. In areas without sufficient observations， the positioning errors are increased by 0.524 m， 0.111 m， and 1.549 m in the North， East， and Up directions respectively， indicating the necessity of using the P₁ and P₂ dual-frequency observation data.

Abstract:In order to reduce the leakage of the water supply network， a coupling modularity optimization and spectral clustering algorithm is proposed to implement pipe network partitioning. First， the modularity optimization algorithm is used in this algorithm to obtain the coarse partition of the maximum modularity in the water supply network. Next， a dual graph is constructed with each partition as a node and the partition connection relationship as an edge. The divisional structure of the pipeline network is affected by the combination of the topology of the pipeline network and the layout of the street， land use， and population distribution. After that， the edge weights are calculated based on the number of points of interest in each district， the average degree of the district， the average pipe diameter， the average pipe length， and the average node elevation. Finally， Laplace matrix decomposition and K-means clustering are used to complete the partition. The experiments verify that the community structure of the pipe network identified by the algorithm is consistent with the actual spatial distribution. A comparison of the algorithm with the modularity optimization and spectral clustering algorithms in the modularity and the number of boundaries indicates that the result obtained by the algorithm is more tightly connected within the partition， the connection between the partitions is sparse， and the boundary pipe is fewer.

Abstract:To study the influence of goodwill impairment decisions on stock price， the equity market reaction of goodwill impairment decisions is examined based on a comprehensive sample of 885 goodwill impairment charges of Chinese listed companies in 2018. The event study method is adopted to calculate the short-term and long-term market performance while the multiple regression methodology is utilized to test influencing factors. The regression findings indicate that short-term market performance of a company is negatively associated to the proportion of goodwill impairment in the balance of goodwill. In addition， the goodwill balance negatively affects the long-term market performance of a company. The empirical evidences suggest that goodwill impairments will lead to the decline of stock prices in the short-term， and the decline rate depends on the proportion of goodwill impairment. The reduction of goodwill balance can positively affect the market performance of companies in the long-term.

Abstract:Most studies have discussed the impact of empowerment leadership on subordinates behaviors， but ignored the impact of subordinates' expectations. So this paper based on implicit leadership theory and expectation theory， discussed the impact of the congruence of “leader empowerment behavior” and “follower empowerment expectation” on followers’ work performance from a dual perspective. Constructed a theoretical mode of work engagement mediates between the congruence of leadership empowerment behavior and followers’ empowerment expectation and work performance， in order to make up for the deficiency of ignoring the role of subordinates. Through polynomial regression and response surface analysis on a survey of 168 pairs of leaders and followers from 6 enterprises in 4 cities， the findings are as follows： ① there are four matching situations between “leader empowerment behavior” and “follower empowerment expectation”. The congruence of leadership empowerment behavior and followers’ empowerment expectation has more positive impact on followers’ work performance than mismatching. ② when the match is consistent， “leader high empowerment-follower high expectation” has more significant positive effect than “leader low empowerment-follower low expectation”. ③ when the matching is not consistent， “leader low empowerment-follower high expectation” has more significant positive effect than “leader high empowerment-follower low expectation” on the followers’ work performance. ④ work engagement mediates between the congruence of leadership empowerment behavior and followers’ empowerment expectation and work performance.