The eightnode and fortyDOF piezoelectric shell element were applied to shape control of a flat shell structure. By the direct and converse effects, a distributed piezoelectric sensor layer was used to monitor the shape deformation and a distributed actuator layer was used to suppresse the deflection. A finite element model was for static response of laminated shell with piezoelectric sensors/actuators was derived. The model was verified by calculating piezoelectric polymeric PVDF bimorph beam. The results are in good agreement with those obtained by theoretical analysis of Tzou\[1\] and Hwang\[2\] . A case study of the static shape control of a flat shell structure is presented.

A modal controller design of defective system was presented,and it was applied in the flutter control problem of an airfoil. The results show that under a given condition, the flutter frequencies may be defective, and the defective modes may be uncontrolled. The uncontrolled modes can not be stabilized using the single input control. In order to stabilize the uncontrolled modes, it is desirable to introduce the multiinput control and to use the multistage design procedure for modal controllers.

Based on the matrix perturbation and Epsilonalgorithm, a new static displacement reanalysis for modified structure was given. The solution step is straightward and it is easy to implement with the general finite element analysis system. The computational effort is much smaller than the effort needed for the full analysis of the modified structures. The epsilonalgorithm was implemented for a chassis structure. By comparing with the exact solutions and the kirsch combined approximate solutions, it is shown that the excellent results are obtained efficiently for very large changes in the design.

The concepts of sensitivity of strain energy density of the element and the nondimensional measure used for identification of the element to be modified were presented, and they were applied in topological optimization of bus ceiling under multiple loading cases. The optimization was implemented for bus ceiling. Results show that the optmization arithmetic is effective.

The uncertainties affecting the fatigue and fracture of structure were regarded as random variables. Combining stochastic finite element method with reliability analysis reliability analysis for a cracked plate from the aspects of probability theory and mathematical statistics was presented. And MonteCarlo simulations were made for fatigue cracks. From the comparisons of the outcomes of stochastic finite element method with MonteCarlo simulation results, this method is proved to be applicable.

The uncertain parameters were described by the interval variables. Using the interval theory, the vibration control problem of structures with interval parameters was discussed, which was approximated by a deterministic system. The independent modal space control was used to obtain the modal gains. A method to solve the interval eigenvalues of the closedloop system was presented. The expressions of the interval stiffness and interval mass matrix were deduced directly from the interval parameters. With matrix perturbation and interval extension theory, the algorithm for estimating the upper and lower bounds of the interval complex eigenvalues was presented. The results were used to estimate the robustness of the stability of the secondorder control systems, and finally the method was apply to a example.

Two elementary topological changes for axial symmetric structures were presented and a set of explicit formulae of elementary topology changes of axial symmetric structures were given by using MoorePenrose inverse theory and a factorization of stiffness matrix. Utilizing these formulae, a new reanalysis method for topological modifications of static structures were presented. This method has advantages of that the repeated assembling and solving of a set of linear equations is not needed, and all the reanalysis can be obtained by the formulae of topological changes. This method is an exact method and has little computational effort. It is suitable for the reanalysis of topological optimization designs especially the local optimal designs because the formulae are easy to realize on the computer. An example was given to illustrate the valid of this method.

To avoid calculating Hessian matrix when the object function and the constraint functions are approximated by the secondorder taylor expansion, a firstorder method called Davidon and FletcherPowell method (DFP method) was used to obtain the inversion of the Hessian matrix, which completely eliminates the need for evaluating the secondorder derivatives and performing matrix inversion. The presented method was implemented for the optimization of a cantilever plate with stiffeners. The caluculation results show that DFP method is effective for structural optimization.

The existing construction control method was improved. The sensitivity of design parameter was presented. By using the sensitivity information, the design parameters and the deflection errors obtained by the computation and the actural measured ones can be adjusted for each cantilever construction segment. So the bridge can be healed propitiously. The present method is simple and feasible for construction control, which can quickly identify and adjust the main parameters for influencing structural state of bridge to reduce the errors and to ensure favourable effectiveness of construction control. The present approach can provide a valuable guide for the construction engineers.

The robustness of responses of vibration control systems with the uncertain parameters was discussed, and the vibration control problem of an uncertain system was approximated by a deterministic one. A method to calculate the upper and lower bounds of responses of the closeloop system with uncertain parameters was presented. The upper bounds of responses can be used to estimate the robustness of responses of the controlled system with uncertain parameters. A numerical example was given to illustrate the applications of the present approach.

In drying process of maize, the mechanism of crack generation was investigated, the distribution of thermo- and hydrostresses inside maize kernel were studied. Employing the general Maxwell model a stress model was proposed, which was then used to simulate the distributions of principal stress and shear stress inside the maize kernel. Simulation results show that both temperature gradient and moisture gradient significantly influence the magnitude and distribution of the stresses. Either principal stress or shear stress which exceeds the yield stress will result in cracking. Results also show that the maximal principal stress occurs inside the cutin endosperm, while the maximal shear stress occurs on the surface of the cutin endosperm.

A new method, rotationassembleaverage method, for the oscillatory boundary layer flow around a circular cylinder, was proposed. This rotation assemble is consist of same systems and same flows except used different lattice grids. By simulating the oscillatory boundary layer flow around a circular cylinder, it was found that fluid particles movement along the direction of getting away the circular cylinder in the parallel direction of oscillatory, the fluid particles movement along to the circular cylinder in the vertical direction, and the second flow phenomenon with four vertexes near the circular cylinder in the same time. The numerical results agree with classical theoretical results. This method can be used to other problems of flows around circular cylinder.

In nonclassical damped linear system a matrix function transformation was introduced to get rid of the damping term which is difficult to solve in the systems. The original differential equation with nonclassical damped term is changed into to a quasi timeinvariant equation with a classical damped one. A matrix which has a set of complete eigenvector system can be similar orthogonally to a diagonal one through reduce it to Jordan’s normal form, and based on this property,the reducing matrix was taken as the transformation one. Thus all eigenvalues and eigenvectors of the system can be obtained. At the same time it is know by analysis about matrix characteristic of the similar orthogonally that not only the characteristic value after the transformation is equal to the ones before the transformation and they are independence to time variable of the system, but also the eigenvectors of the formed system can showed the effect of the time variable on original system. The transformation bears clear physical meaning to nonclassical damped linear systems.

Facing the relevant criterion is not established about the accurate measurement of bonding strength between carbon fiber and concrete, on the basis of the principle of the static force equation, the quantitative relationship between the tensile strain in carbon fiber and the shearing strength of the concrete was derived under the tensile status ultimately. The basic mechanical variable in the theoretical formula is measured by experiment. The function relation of strain along the length of the specimen can be gotten through function fitting. Then the distribution rule of shearing strength and shear intensity can also be obtained. The reliable design parameters can be obtained by the presented method for the design of the carbon fiber reinforced concrete beams.

By using the convex model theory, the vibration control problem of structures with uncertain parameters is discussed, which is approximated by a deterministic one. The uncertain parameters are modeled to be a convex elliptical set rather than a probabilistic set. This does not require the probabilistic distribution descriptions of the uncertain parameters. The feedback gain matrices derived from the deterministic systems are applied to the actual uncertain systems, and a method for estimating the upper and lower bounds of the real and imaginary parts of eigenvalues of the closedloop systems is presented by combining the secondorder matrix perturbation and optimization. The numerical results show that the present method is effective.

A dam breaking wave is similar to the shock wave that is a solution of gas dynamic equations. Based on the analogy to gas dynamic equation, the KFVS method can also be used to stimulate the shallow water wave equations. Two numerical tests were calculated. The numerical results agree with those of available literatures and show that the method is robust and stable for getting the discontinuous solutions of the shallow water wave equation.

Abstract：There are many problems in the carling forming and unloading processes such as expand and shrink of U section wall, the torsion of U section and the warp of carling head. These problems are concerned by the carling factories and do not solved for a long time. The mechanism of the carling forming and spring back is analyzed. The carling forming and spring back processes are simulated by KMAS software. The warp of carling underside is researched by half analytic and half experiment method and the best scheme is found by this method. The new die is on stream and the best effect is obtained. The reliability and practicality value of the research method are proved.

To manufacture the intensive and complicated automobile panel quickly and accurately, the forming process was studied. The spring back in the distortion process was predicted by finite element method (FEM), and the predicted results were analyzed to find out the reason lead to spring back, and so as to precompensate the spring back. A new compensation method for spring back was established by FEM, and the validity was verified through the stamping experiment of practical part, thus providing a theoretical basis for the proper forming process of automobile panel.

Key techniques about how to smooth the contour of automotive panel were discussed. The method for judging the areas which need to be smoothed and the procedure used for generating the NURBS smooth curve of the stamping part contour were presented. The relevant function modules for parametric design were developed. The feasibility of techniques was proved by creating the smooth surface on the contour of practical parts.

Nanograde NiO/Ce0.8Gd0.2O1.9 compound powders were synthesized via citrate/nitrate solgel method. TEM and XRD results show that uniform and ultrafine double phase compound powders can be obtained by this method. Analysis of the veaction process and the powder show that the molar ratio of citric acid to metal ions (MRCM) in the precursors is vital to get the desirable powders. When the MRCM is equal to 1.5 the target powders with 13~18 nm crystalline size are produced.

The design of the electronic control unit for the metal belt continuously variable transmission , including hardware integration and software compilation based on Infenion C164CI was presented. Design principle of the input module and the output module was investigated. Utilizing the modularizing design concept and the of realtime and multitask control technology, the realtime optimized management on the control system was realized. To validate the design, a bench test was performed and the results show that the electronic control unit fulfils the control requirements satisfactorily.

A new type of energysaving assembly—an overrunningsliding clutch on the end of the vehicle driving axle shaft was introduced, its structure and working principle were discussed, and its performance tests were done. A mathematical model which integrates the internal energy dissipation of the transmission system and the external resistance energy consumption was built from the view point of the internal energy dissipation and momentum, and verified by the road tests. Results showed that the introduced new type device is characterized by significant energy saving effect with wide application prospects.

An algorithm for collision detection in threedimensional (3D) lathingsimulation was developed based on the study of current collision detection methods. In modeling process, the polygon concept and B_rep molding method were employed. The algorithm converts 3D solid model into 2D polygon model, and the occurrence of collision is detected by identifying the intersections among the 2D polygons using Boolean operation. This detection method can be performed on the 3D dynamic lathing-simulation system developed by the authors. Case study verified that this method can simplify the collision model, hence accelerates the detection process.

To accomplish the wide range and high precision speed measurement in the attitude control and energy storage flywheel application, a new measurement method was put forward based on the analysis of the errors. In the low speed region, the time of several integrated flywheel speed pulse signal was measured. While the speed exceeds a certain threshold speed, the pulse number in a certain time was counted. For proving the feasibility of this method, the speed measurement experiment was done by using high precision arbitrary waveform generator and universal counter produced by Agilent Company, and the experiment results showed that the suggested speed measurement method can meet the demand of the integrated flywheel energy storage and attitude control system with the advantages of high precision and wide range.

In order to study the size effect of the axialmagnetized permanent magnet dual rotor micromotor on its performance, the elementedge based finite element analysis method was adopted to simulate the torque of the motor. The simulated torques of a motor with 20 mm diameter were compared with the experimentally measured ones and a good agreement between them was achineved. The winding density in the stator was analyzed, and the obtained optimal scheme is that both the line width and the space between the lines are 50 μm. The minimum dimension of the motor confined by the configuration, torque of 47 μN·m, and air gap magnetic flux density, etc., was also analyzed and it was reached that the motor with 10 mm diameter and 5.2 mm thickness may satisfy the practical application needs in the area of fine mechanics with the electric current of 0.1A. The suggested analysis method is beneficial to the design and miniaturization of the micromotor.

The design, fabrication and evaluation of a new microchamber electromagnetic mixer were described. The dynamic analysis was applied to the mixer in order to improve its mixing performance. The computational results proved that the mixing efficiency is dependent on both frequency and amplitude of membrane vibration. The polymethylsiloxane(PDMS) mixers used in optical evaluation were fabricated by mold pressing. The planar coils manufactured by DELIGA are characterized by compactness and good adhesion. The hybrid integrated mixer chip consists of PDMS mixer, magnet, optical lens, planar coils and mini spectrometer, et al. Its mixing experiments show that the 6 mm mixer with membrane thickness 50 μm achieves good mixing of volume 150 μL in 5 s.

Based on the analysis of the limitations of the typical methods to detect or extract the signals in the chaotic and noisy background, a new method was proposed. The statistical independence of the signals measured by the negentropy in the information theory was used to distinguish the chaos and the signals, the independent component analysis technique was used to separate progressively the signals from the chaos and noise, and thus the signals were detected. The computer simulation showed that the proposed method can detect not only the strong signals but also the weak signals in the powerful chaos and noise with great stability and reliability.

The pairing steps were always needed to estimate the frequency pairs of harmonics which was described by the complex signals. The quaternion signal model was constructed through the separated signal and its Hilbert transform. The corresponding four order cumulant was defined to restrain the additive Gauss noise and the frequency was extracted using the properties of Hamilton quaternion and hypercomplex. Using this method there is no need to construct the complicated extended matrix, and the error frequency pairs and the imbalance of the two frequencies of the frequency pairs may be accounted for. The algorithm was verified by simulations.

The variance modeling of product line was realized through the extension mechanism of UML Profile, and implements the general constraints of the product line model and specific constraints of the concrete product line model using UML OCL on models of metamodel layer and model layer seperately, and gives the algorithm of product model library derived from product line model.

In order to solve the problem that the traditional single threshold method can not segment the shadow path mark image successfully, a new algorithm based on Regiongrowing has been proposed for our indigenously designed and manufactured vision navigation intelligent vehicle (Cyber Car). It utilizes the path mark edge to search several kernel points. Therefrom the path mark region grows and the segmentation and extraction of the navigation path mark can be achieved. The divided subimages compression method was proposed to fulfill the realtime demands. The segmentation experiment results of different shadow path mark images show that the new method can segment the navigation path accurately and quickly with good antijamming ability.

Based on analysis of the calibration method of threedimensioned photogrammetry, a new method of camera calibration using planar patterns was proposed.After several photos were obtained from different views, a linear solution was derived, followed by a nonlinear refinement based on the maximum likelihood criterion, the final parameters of the model were achieved. The experiment results demonstrate the proposed method is not only simple and accurate, but also flexible.

The realtime traffic information collection and transmission in the integration of the urban traffic flow guidance system(UTFGS) and the urban traffic control system(UTCS) were discussed and a theoretical model for the information processing in the integration of these two systems was described. It was suggested that a traveler behavior based dynamic traffic assignment(DTA) be applied to the processing and a DTA model for the optimization of the global route was presented. A computer supported cooperative network was applied into the network communication in the integration of UTFGS and UTCS and a data communication based on the multicast communication in the integration of two systems was given.

The merging ramp metering and main road dynamic speed guidance technique for the effective management on the urban expressway was introduced and their coordination strategy of control and guidance was suggested. A traffic flow model of the expressway under the speed guidance condition was derived and an optimal coordination function for ramp metering and main road speed guidance to maximize the expressway service capacity and minimize the ramp queue delay was built. The suggested coordination strategy was simulated by the software VISSIM. The results showed that with the dynamic speed guidance on the main orad and the ramp metering the urban expressway traffic flow capacity increases over 10%, and the average travel time decreases over 30%.

A simplified traffic flow model for the road intersection was proposed and a fuzzy logic controller based on the phase sequencer was developed. The developed controller was designed based on the simulation of the traffic policemen judging and decisionmaking process during their manual traffic control. A new algorithm was used which is derived from the system concept and differs from the traditional signal coordinated control algorithm. The fuzzy controller and the phase sequencer are combined to perform the traffic control, the former decides the desirability of phase switch, and the latter decides the next green phase based on the priority rating. The computer simulation showed that the developed controller works with good performances.

The factors affecting the urban traffic system were discussed and concluded that the urban traffic problem should be solved by the crossing and merging of the different disciplines to form a comprehensive solution. The causes to make the urban traffic system very complex were revealed, and its research method was proposed. The servo parameters of the urban traffic complexity system were found, the steps of researching the urban traffic system by the macroscopical method and the design concept of the urban traffic management decisionmaking system by means of the complexity theory were formulated.

Abstract：A traffic flow assignment model for the coordination optimization between the traffic flow guidance and the traffic control was built with minimizing the integrated weighted congestion rate of both the road section and the intersection as the optimization target, and the genetic algorithm was applied to solve the model. The adjustment to equilibrate the road network traffic flow from the coordination between the guidance and control obtained from the realtime rolling computing the model is a nice quantitative support to provide the guidance information for the vehicle drivers to aooid traffic congestion. The test on a small road network proved the algorithm of the model is characterized by good realtimeness, high effectiveness and easy implementation.