OPTIMAL PWM BASED ON REAL—TIME SOLUTION WITH NEURAL NETWORK

A novel concept of neural network based control in pulse-width modulation(PWM)voltage source inverters is presented.On the one hand,the optimal switching an-gles are obtained in real time by the neural network based controller；on the other hand,the output voltage is ad-justed to fit the expected value by neural network when input voltage or loads change.The structure of neural network is simple and easy to be realized by DSP hard-ware system.No large memory used for the existing opti-mal PWM schemes is required in the system.Theoreticalanlysis of the proposed so-called sparse neural network is provided,and the stability of the system is proved.Un-der the control of neural network the error of output volt-age descends sharply,and the system outputs ac voltage with high precision.     

OPTIMIZATION BASED ON LMPROVED REAL—CODED GENETIC ALGORITHM

An improved real-coded genetic algorithm is pro-posed for global optimization of functionsl.The new algo-rithm is based om the judgement of the searching perfor-mance of basic real-coded genetic algorithm.The opera-tions of basic real-coded genetic algorithm are briefly dis-cussed and selected.A kind of chaos sequence is described in detail and added in the new algorithm ad a disturbance factor.The strategy of field partition is also used to im-prove the strcture of the new algorithm.Numerical ex-periment shows that the mew genetic algorithm can find the global optimum of complex funtions with satistaiting precision.     

Effect of Particle Size Distribution on Radiative Heat Transfer in High-Temperature Homogeneous Gas-Particle Mixtures

The weighted-sum-of-gray-gas(WSGG)model and Mie theory are applied to study the influents of particle size on the radiative transfer in high temperature homogeneous gas-particle mixtures,such as the flame in aero-engine combustor. The radiative transfer equation is solved by the finite volume method. The particle size is assumed to obey uniform distribution and logarithmic normal(L-N)distribution,respectively. Results reveal that when particle size obeys uniform distribution,increasing particle size with total particle volume fraction fvunchanged will result in the decreasing of the absolute value of radiative heat transfer properties,and the effect of ignoring particle scattering will also be weakened. Opposite conclusions can be obtained when total particle number concentration N_0 is unchanged.Moreover,if particle size obeys L-N distribution,increasing the narrowness index σ or decreasing the characteristic diameter Dˉwith the total particle volume fraction fvunchanged will increase the absolute value of radiative heat transfer properties. With total particle number concentration N_0 unchanged,opposite conclusions for radiative heat source and incident radiation terms can be obtained except for radiative heat flux term. As a whole,the effects of particle size on the radiative heat transfer in the high-temperature homogeneous gas-particle mixtures are complicated,and the particle scattering cannot be ignoring just according to the particle size.     

CALCULATION FOR KERNEL OF INTERVAL GREY NUMBER BASED ON BARYCENTER APPROACH

The kernel of interval grey number is most likely the real number,which can be used to represent whitenization value of interval grey number.A novel method for calculating kernel of interval grey number is constructed based on the geometric barycenter of whitenization weight function in the two-dimensional coordinate plane,and the calculation of kernel is converted to the calculation of barycenter in geometric figures.The method fully considers the effect of all information contained in whitenization weight function on the calculation result of kernel,and is the extension and perfection of the existing methods in the scope of application.     

MAC Models in Thermoelasticity

There are two types of singularities in the linear thermoelasticity.The first one arises in the field of stresses if a force is applied to one point of the body.This singularity is physical and should be accepted.The second type of singularities is nonphysical and they arise in the fields of displacements and temperatures.There exist the nonlocal theories and gradient theories which have the goal to introduce the finite stresses instead of the infinite ones.The MAC model of the thermoelasticity is created to avoid the nonphysical singularities and it accepts the infinite stresses.MAC is the method of additional conditions,which allows introducing the new model to use the classical model,plus additional condition of the physical nonsingularity and/or condition of the good behavior of the solutions at infinity.The MAC Green′s functions for the heat conduction and for the elasticity could be introduced using the differential MAC models.The infinite and finite bodies are considered.The principle of superposition is applied to obtain the integral equations to solve the boundary value problems.The strength criteria based on finite stresses could be changed in this model because the infinite stresses are allowed.The strength criteria based on deformations are applicable.Classification of MAC models is given.     

快速可重构信息系统的框架研究

Changes and reconfiguration of information systems(ISs)are two kernel concepts of rapidil reconfigurable in-formation system(RRIS).It is the abilitise to evolve with changes that make RRIS superior to the traditional ISs.Hierarchical architecture of RRIS is put forward.And then a component-based framework of RRIS is discussed including its building and designing in detail.     

NUMERICAL AND EXPERIMENTAL STUDY OF LURE MOVING IN WATER

The motion of a lure in water is investigated experimentally and numerically. The lure motion in water of a passing water tank is observed, and the periodic motion is/ound. From the Fourier analysis, it is found that the frequency with the largest amplitude in the lateral direction depends on the lip width of the lure. To understand the lure dynamics, a numerical simulation of the flow field around the lure is performed. The shape is measured using an X-ray computer tomography and converted into a voxel model. From visualization, it is found that vortex sheds from its lip correspond to the vibration frequency of the lure.     

激光选区烧结金属粉末的研究

A series of experiments are carried out with selective laser sintering(SLS) for copper,copper-nickel powder matetials.The phenomena of the sintering process is ana-lyzed,the influence of techmological parameters is dis-cussed in detail,By means of SEM with energy spectrum system and X-ray diffraction spectrum,the morphology of the microstructure and the compositions of the different zones are analyzed,the mechanism of the laser sintering metallic powders is preliminarily explored,All those above lay the foundations for the forming metal parts by laser sintering metallic powders.     

A Novel 2D Piezo-Nanopositioning Stage Based on Triangle Amplifier Mechanism

To satisfy the demand on dynamic performance and load characteristics of piezoelectric actuators in aeronautics and astronautics fields,a novel 2Dpiezo-nanopositioning stage utilizing a triangle amplifier mechanism is proposed.The stage is driven by piezoelectric rhombic units in both X and Ydirections,which is composed of four piezoelectric stacks.Theoretical static model develops the relationships among output force,displacement,static stiffness and the structure parameters of the platform.The experimental results of the prototype show that the output performances in X and Ydirections are similar and both of them are within an 8% deviation from the theoretical values.The stroke of the stage reaches 41.6μm and 42.9μm in Xand Ydirections,respectively,and is directly proportional to the amplitude of the input sinusoidal voltage 10 Hz.Moreover,the nano-positioning stage is featured with bidirectional symmetrical output characteristic and millisecond starting characteristic,whose minimum output displacement step is 50 nm.     

战斗机的体系对抗分析

Analyses of the systematic confrontation between two military forcfes are the highest hierarchy on opera-tional effectiveness study of weapon systema.The physi-cal model for tactical many-on-many engagements of an aerial warfare with heterogeneous figher aircraft is estab-lished.On the basis of Lanchester multivariate equations of square law,a mathematical model corresponding to the established physical model is given.A superiorityh parame-ter is then derived directly from the mathematical model.With view to the high -tech condition of modern war-fare,the concept of superiority parameter which more well and truly reflects the essential of an air-to-air en-gagement is further formulated.The attrition coeffi-cients,which are key to the differential equations,are de-termined by using tactics of random target assignment and air-to-air capability index of the fighter aircraft.Hereby,taking the mathematical model and superiority parameter as cores,calculations amd analyses of complicate systemic problems such as evaluation of battle superiority,prog-mostication of combat process and optimization of colloca-tions have been accomplished.Results indicate that a clas-sical combat theory with its certain recent development has received newer applications in the military operation research for complicated confrontation analysis issues.     

FREQUENCY LOCK—IN PHENOMENON OF FLAPPING MOTION WITH SURFACE WATER WAVE

To study the effect of a progressive surface wave on the flapping motion of a vertical turbulent plane jet in shallow water,the laser visualizations and measurements were carried out in a water tank.The images of visualization by laser-induced fluorescence(LIF)technique show that the jet flapping motion occurs in the wave environ-ment.Using the wave height gauge and laser Doppler velocimetry(LDV) ,experimental results show that the jet flapping motions indeed lock-in to the wave oscillations if the wave frequency is close to but lower than the nature frequency of jet flapping motion.The phenomenon does not occur when the wave frequency is above the nature frequency of jet flapping motion.     

MINING CYCLIC GENERALIZED ASSOCIATION RULES

Discovering cyclic generalized association rules from transaction datbases can reveal the relationship of differ-ent levels of the taxonomies and display cyclic variations over time.Information about such variations is great use of better identifying trends in associations and forecast-ing.Because cyclic rules are quite sensitive to a littlenoise,this paper uses the noise-ratio as the criterion of i-dentifing cydclic itemsets for dealing with the problem and utilizes the cycle-pruning technique to reduce the comput-ing time of the data mining process by exploiting the real-tionship between the cycle and generalized frequent item-sets.The paper gives the algorithm of mining cyclic gen-eralized itemsets(CGI).Experiment shows that the CGI algorithm can efficiently yield results.     

RESEARCH ON SCHEME OF HIGH-SPEED ROTOR／WING TRANSITION HELICOPTER RD15

To analyze the existing schemes of high-speed rotorcrafts and some new technologies, a new conceptual sketch of the high-speed rotor/wing transition helicopter RD15 is proposed. The overall layout of the RD15 is given out and the transition process from the helicopter mode to the airplane mode is designed. The lift system consists of a circular disk-wing with four retractable blades. The technology of individual blade control is adopted for flight control in hover and low speed flight. The tail is a vectored thrust duct propeller. It can provide the anti-torque in hover, and offer the multi-directional controls and propulsion drive for the airplane mode flight. The aerodynamic characteristics and key technologies in the transition process for this layout, including the nose up angle of disk-wing, the length of the blade, rotation speed, pitch angle and other parameters, are theoretically ana lyzed and experimentally tested. Calculation and experiments show that the shift process of the lift, the power and controls are smooth, and the designed scheme is feasible.     

Simulation of Air Source Heat Pump with Refrigerant Injection for Heating Operation

An air source heat pump(ASHP)with refrigerant injection is proposed for the air conditioning system of electric vehicles(EVs),especially for efficient heating in cold winter,when there is no wasted heat of engines.The simulation model is built with the framework of two-phase fluid network,where the compressor is separated as two compressors and the economizer is treated as two heat exchangers in the injection path and the main refrigerant path.With the validated simulation model,the heating performance is analyzed,and the results show that the coefficient of performance(COP)of ASHP with refrigerant injection is higher than 1.4 and the discharge temperature is less than 100℃ when the outdoor temperature is-20℃.The above performance ensures that the air conditioning system and EVs can operate normally with high efficiency even in the cold winter,which is much helpful for the practicability of EVs.     

Fractional Pfaff-Birkhoff Principle and Birkhoff′s Equations in Terms of Riesz Fractional Derivatives

The dynamical and physical behavior of a complex system can be more accurately described by using the fractional model.With the successful use of fractional calculus in many areas of science and engineering,it is necessary to extend the classical theories and methods of analytical mechanics to the fractional dynamic system.Birkhoffian mechanics is a natural generalization of Hamiltonian mechanics,and its core is the Pfaff-Birkhoff principle and Birkhoff′s equations.The study on the Birkhoffian mechanics is an important developmental direction of modern analytical mechanics.Here,the fractional Pfaff-Birkhoff variational problem is presented and studied.The definitions of fractional derivatives,the formulae for integration by parts and some other preliminaries are firstly given.Secondly,the fractional Pfaff-Birkhoff principle and the fractional Birkhoff′s equations in terms of RieszRiemann-Liouville fractional derivatives and Riesz-Caputo fractional derivatives are presented respectively.Finally,an example is given to illustrate the application of the results.     

Crack Detection in Pipes with Different Bend Angles Based on Ultrasonic Guided Wave

Pipeline plays an indispensable role in process industries,because the progressing crack-like defects of in it may result in serious accidents and significant economic losses.Therefore,it is essential to detect the cracks occurred in pipelines.The axial crack-like defects in elbows with different angle are inspected by using the T(0,1)mode guided waves,in which different configurations including 45°,90°,135°and 180°(straight pipe)are considered respectively.Firstly,the detection sensitivity for different defect location is experimentally investigated.After that,finite element simulation is used to explore the propagation behaviors of T(0,1)mode in different bend structures.Simulation and experiment results show that the crack in different areas of the elbow can affect the detection sensitivity.It can be found that the detection sensitivity of crack in the middle area of the elbow is higher compared to the extrados and intrados of the elbow.Finally,the mode conversion is also investigated when the T(0,1)crosses the bend,and the results show that bend is a key factor to the mode conversion phenomenon which presents between the T(0,1)mode and F(1,2)mode.     

Multi-factor Effects on Layout of Solar Collector

Considering the influencing factors of the layout of solar collectors such as the tilt angles,azimuth angles,spacing between collectors and the number of collector rows,a mathematical model of the collected energy of the solar collector with limited area on the horizontal plane is established. Two different optimized models including the cost?benefit model and the minimum annual auxiliary heating energy model are conducted in this paper. The results show that,the collected energy in a year could increase with the increase of the number of collector rows. And the collected solar radiation in a year increases firstly and then decreases with the increase of the collector tilt angles.Furthermore,the collected solar radiation in a year increases firstly and then decreases with the continuous increase of the azimuth angles from-90° to 90°. Taking Nanjing city of China as an example,based on the optimized objective of maximum benefit,the optimal layout of the solar collector array in the area of 200 m2 should include:the number of collector rows is 8,the tilt angle is 40° and the azimuth angle is 0°. Meanwhile,the optimal methods for the optimized objective of minimum annual auxiliary heating energy should include:the number of collector rows is 9,the tilt angle is 50°,and the azimuth angle is 0°.     

Modeling and Simulation of Pipeline Security Monitoring System Based on Line-Structure Sagnac Interferometer with 3×3Coupler

The third damage action,human disruption behavior,is one of the major threats to pipeline operation.It is necessary to monitor and locate the perturbation behavior that may threat pipeline safety in real-time.Therefore,a new pipeline security monitoring system is designed by using line-structure Sagnac optic fiber interferemeter with the characteristic of 3×3coupler that can modulate the phase of optic signal,with no need for phase modulation and demodulation.The optic structure of the new system is simplified,the signal processing accuracy is improved,and the polarization effect is reduced.The working principle of the monitoring in ideal condition is described,the phase demodulation is analyzed and the location of the possible damage action point is formulated.By using simulink,the optic signal propagation and interference is simulated and interference signals in different frequencies are obtained.The results validate the feasibility of the monitoring system and indicate that the low frequency signal less than 1kHz resulting from the human damage action can be detected.The disturbance set at 10 km can be located by calculating the demodulation signal accurately over a long monitoring distance.     

Compensation Algorithm Based on Estimation State for Transmission Delay in Cooperative Localization

In order to maximize the utilization of the observation information in the cooperative localization, a com- pensation algorithm based on the estimation state is presented for transmission delay. Under the framework of the Kalman filter, two different processes of state estimating with and without transmission delay are investigated and contrasted. The expression of difference quantity caused by transmission delay is derived. It is used to compensate the present estimation state instead of the observed information compensation. According to the characteristics of state transition matrix, an equivalent expression of which successively impacts on the covariance factor in delay time is obtained. The simulation results show that the present estimated state is effectively corrected by transmis- sion information and the relevance among agents is accurately updated. As a result, a higher positioning accuracy is achieved. Meanwhile, the consumption of recording and multiplication of the state transition matrix is saved.