In order to solve the problem of real time reactive power compensation, a reactive power compensation method based on double random phase estimation was proposed. The Fourier transform was carried out for the sampled power signal, and the fundamental wave of power signal and the frequency information of harmonic waves in each order were determined with the spectral peak location algorithm. In addition, the signal with the same frequency and different phase was modulated to obtain the modulated signal, and the filtered signal was attained through the low pass filtering. In combination with the specific value of filtered signal and the value of random phase trigonometric function, the signal phase information to be estimated was obtained, which could assist the reactive power compensation. The validity of the proposed algorithm was verified through the stimulated and the actually acquired power signals. The results show that the relative error of phase estimation is below 1‰, and the reactive power compensation can effectively improve the energy usage efficiency of the circuit.

Aiming at the problem that the photovoltaic grid-connected inverter in conventional constant power control mode lacks the dynamic adjustment ability of voltage and frequency, a control strategy of photovoltaic grid-connected inverter based on virtual synchronous generator(VSG)was proposed. According to the principle of synchronous generator, the grid-connected equivalent circuit and vector relation expression of synchronous generator were established, the active frequency control algorithm and reactive voltage control algorithm were designed, and the photovoltaic power generation system based on the control strategy of VSG was established. The 10 kW photovoltaic grid-connected system was established in Matlab/simulink environment. The simulated results show that the photovoltaic grid-connected inverter based on VSG has the similar frequency and voltage regulation characteristics with the synchronous generator, and can well adapt the requirements of grid operation.

In order to investigate the cyclic stress response behavior and low-cycle fatigue behavior of Incoloy 825 nickel-based alloy, the applied total strain amplitude controlled fatigue tests at both room temperature and 760 ℃ were carried out, and the strain fatigue parameters of the alloy at different temperatures were determined. The dislocation substructures and morphologies of fracture surfaces for the alloy after low-cycle fatigue deformation were analyzed with the transmission electron microscope(TEM)and scanning electron microscope(SEM). The results show that at room temperature, the alloy exhibits the characteristics of cyclic hardening followed by the cyclic softening, while the alloy exhibits the characteristics of cyclic hardening followed by either cyclic softening or cyclic stability at 760 ℃. The relationship between both plastic and elastic strain amplitudes as well as reversals to failure at different temperatures for the alloy shows a single slope linear behavior, which can be described by the Coffin-Manson and Basquin equations, respectively. At both room temperature and 760 ℃, the low-cycle fatigue deformation mechanism for the alloy is mainly the planar slip. In addition, the fatigue cracks initiate and propagate in the transgranular mode.

In order to determine the effect of melt superheating and annealing temperature on the magnetic properties of Fe₇₈Si₉B₁₃amorphous ribbon, the amorphous ribbon with a width of 142 mm and a thickness of 23 μm were prepared with the melt-spun method at three different superheating temperatures of 180, 200 and 230 ℃, respectively. The annealing treatment for the amorphous ribbon was performed in the longitudinal magnetic field environment at five different temperatures from 350 to 370 ℃, and the magnetic properties of the annealed amorphous ribbon were tested with the silicon steel measuring instrument. The results show that with increasing the melt superheating, the exciting power and core loss of amorphous ribbon tend to decrease. With decreasing the superheating and annealing temperature, the distribution region for the magnetic properties of amorphous ribbon becomes narrow while the magnetic stability of amorphous ribbon increases. It is helpful in obtaining the amorphous ribbon with the optimal magnetic properties to increase appropriately the superheating and heat treatment temperature.

In order to improve the wettability between carbon nanotube(CNT)and Fe substrate, a N-doped CNT supperlattice structure with finite length was constructed. The energy calculation results based on the first principle show that the embedding energy of novel superlattice structure positively increases, the structure stability decreases, but the Fe atom adsorbing ability of outside wall gets significantly improved. The difference charge density indicates that in the doping system, the π bond connecting N atom and neighboring C atom has distorted, which allows N atom to bond with Fe atom easily. Both population and charge transfer situation show that the electron loss ability of Fe atom decreases due to the doping of N atom, but the intensity of Fe—N covalent bond increases. The superlattice structure can hold the ability to adsorb Fe atom under some extent of torsional and shear deformation.

Aiming at the fatigue failure problem of tension shaft for a turbofan engine in the working environment, the theory model based on the combination between finite element method(FEM)and the Miner linear cumulative damage theory was proposed. The calculation of standard cycle load spectrum was implemented under the condition of knowing the materials and geometry size of tension shaft. Furthermore, the fatigue life test and analysis for the fan tension shaft were carried out. The results show that in the test, the tension shaft can bear 30 000 times of test cycle load, and the tension shaft does not exist such failure phenomena as the cracks, cutting of convex shoulder and broken buckle of thread after the test. The security service life was determined under the standard cycle load. Through performing the fatigue test research and simulation calculation analysis for the tension shaft, the feasibility of both reliability life prediction method and model is verified.

Aiming at the problem that the extended bonhoeffer-van der pol oscillator is easily affected by uncertainty factors, a method in combination with both nonlinear dynamics analysis and numerical simulation was proposed, and the influence of weak perturbation on the mixed-mode oscillation type and synchronization behavior was investigated. According to the simulated results, it is noted that the collapse phenomenon appears because the mixed-mode oscillation in several types is affected by the weak perturbation. It is theoretically proved that the coupled strength needed to reach the complete synchronization for the membrane voltage of two coupled bonhoeffer-van der pol oscillators is larger than that for the slow variable. On a 2D-parameter plane, the parameter range for the membrane voltage to reach the complete synchronization should be smaller. The results indicate that compared with the membrane voltage, the slow variable is easier to achieve the synchronization. Furthermore, a linear relationship between the variables is presented.

In order to optimize the acoustic computer tomography(CT)reconstruction system for the 3D temperature field, the reasonable selection of sound path was performed with simulation research. Twenty sound transceivers were arranged around a cuboid measurement area. Four acoustic path selection modes were considered, all non-duplicated paths were retained, and the paths on the edges, sides and surfaces of the cuboid were removed. In addition, the relative error and root-mean-square error were used to evaluate the reconstruction quality of temperature field. The results show that the selection modes of sound path have an obvious effect on the reconstruction of temperature field. The errors and computation amount in the acoustic CT method for the reconstruction of temperature field can be reduced through the reasonable selection of sound path.

Aiming at the low accuracy and poor efficiency problems when the traditional mining methods are applied to the abnormal data mining in the remote virtual education communication, an abnormal data mining method based on FWSCA and differential evolution method in the remote virtual education was proposed. The data characteristics of remote virtual education communication were extracted with the information gain method. In addition, the data characteristics of online communication were clustered with the introduction of WTA rule. On this basis, the data were distinguished with the sparse score method, and the FWSCA in combination with the differential evolution method was adopted to conduct the abnormal data mining in the remote virtual education communication. The results show that when it is used for data mining, the proposed method exhibits higher mining precision and short mining time, and has certain advantages compared with the traditional mining algorithm.

Aiming at the problem that the solar energy automatic tracking system has low tracking precision and poor stability in the traditional filtering algorithm because the isolated noise points in the sun light spot can not be completely eliminated and the image edge of sun light spot is fuzzy, a hybrid filtering noise algorithm based on the improved Wiener filtering and the detection mid-value of extreme value was proposed to carry out the denoising and filtering of sun light spot.Through measuring the solar altitude angle and azimuth angle at a certain place in different cases, the simulation experiments were performed.The results show that the proposed algorithm can adapt to the complex weather conditions, and can still conduct the accurate tracking of the sun under the rain and haze weather.Therefore, the proposed algorithm can reach the desired effect of improving the solar utilization ratio.

In order to reduce the effect of environment temperature on the output optical power and center wavelength of near-infrared distrubuted feedback(DFB)laser, a temperature control system of DFB laser was designed with the PID control scheme. The PID control system could obtain the optimium point between the dynamic balance setting time and maximium oscillation amplitude through adjusting the proportion(K_p), integration(K_i), differentiation(K_d)parameters with the hardware circuit. The results show that the control precision of temperature control sysytem is ±0.05 ℃, and the temperature control scope is from 10 ℃ to 50 ℃. After longer term supervision, the temperature of DFB laser maintains a very stable status, and the output center wavelength has no drift, which can satisfy the requirements of high detection accuracy for the gas concentration.

In order to achieve the accurate control for the temperature of diode laser, a high precision temperature control system combining PID control and genetic algorithm was designed. The temperature value of diode laser was measured and fed back with the thermistor and photodiode in the system. Meanwhile, thermo electric cooler(TEC)was used as the actuator to control the temperature. Therefore, a double close-loop control system was constructed. The method of optimizing PID parameters in an on-line way with genetic algorithm was proposed. According to the real-time measured error, controller output and rising time, a function was established as the performance index to be optimized. And thus, the parameters for the controller could be dynamically adjusted. The results show that the system can achieve the accurate and stabilized temperature control. The stability and temperature error are 0.027% and 0.002 ℃, respectively. Furthermore, the control effect is obviously superior to that of traditional PID control. Therefore, the better application value can be obtained.

In order to study the influence of size effect on the mechanical properties of GFRP-concrete-steel tubular composite columns, the mechanical properties of 8 specimens with different specimen sizes and different concrete strength grades under axial compression were tested and studied with the static loading method under the condition of satisfying the geometrical similarity. In addition, the failure phenomenon, load-strain curve, ultimate bearing capacity and stress value change of 8 specimens of composite columns were analyzed and compared. The results show that with increasing the specimen size, the deformation resistance of specimens decreases, and the size effect is obvious. With increasing the thickness of GFRP tube, the bearing capacity of specimens gets improved, the mechanical properties of specimens get enhanced, and the effect of size effect reduces.

The size and shape of skeleton particles in the mixed rock and soil medium have different influence on the mechanical properties of the medium. Based on the 3D laser scanning data of typical geotechnical particles including the pebbles and gravels, the statistical analysis on the shape characterization parameters of the particles was performed with the spheric harmonic function, and the influence change among various statistical magnitudes was dicussed. The results show that for the gravels and pebbles with the same volume, the smaller the volume is, the smaller the difference in the surface area of gravels and pebbles is, and the smaller the influencing area for the mechanical properties is. The overall shape coefficient of pebbles is greater than that of gravels. In addition, the edge degree and sphericity obey a certain linear rule, and share an inverse relationship with each other. The bigger the size of particles is, the obvious the detail exhibition of the particle surfaces is. Furthermore, the details of the surfaces have important influence on the relevant mechanical properties. The shape coefficient can reflect the sphericity degree of the particles to a certain extent. The study results can provide the references for the macroscopic role analysis of rock and soil particles in the relevant mechanical properties.