We considered a class of compressible nonNewtonian fluid equations in onedimensional bounded interval. Under the conditions that the external force term was relatively small, a priori estimates were given by the method of hypothesis closure, weighting and energy estimates. We proved the existence and uniqueness of the global strong solutions and gave the long time behavior of the solution.

We considered the first boundary value problems for a class of steady NonNewtonian micropolar fluid equations in a smooth bounded domain Ω∈Rⁿ (n=2,3). Under the conditions that the vortex viscosity coefficient and a certain norm of the external force term were appropriately small, we proved the existence and uniqueness of strong solutions of the system of equations by using the fixed point theorem with index p>1.

The Hessian matrix in a quadratic model was approximated by a Hessian matrix satisfying the modified secant equation, and the search direction was derived by minimizing the quadratic model in a threedimensional subspace. Combined with some nonmonotonic line search strategies and restart techniques, we proposed an improved threedimensional subspace minimization conjugate gradient algorithm. Under some reasonable assumptions, the global convergence of the algorithm was proved. For the test function set named Andrei, numerical experiment verified the effectiveness of the new algorithm.

Based on a class of orthogonal polynomials, alternative Legendre polynomials (ALPs), we proposed a numerical calculation method for fractional pantograph delay differential equations. Firstly, we obtained the numerical approximation results of fractional calculus by using the properties of ALPs. Secondly, we  transformed the fractional pantograph delay differential equations into the  algebraic system for obtaining solution. Thirdly, we analyzed  the error  of the method  and obtained the convergence results of the method. Finally, some numerical examples were given to verify the effectiveness and accuracy of the proposed method.

We considered the  control problem of the  backward doubly stochastic  system with time delays in  both state variables and the  control variables. By introducing the anticipated backward doubly stochastic differential equati
ons as its adjoint equation,  using the classical convex variational method and dual technique, we gave the  maximum principle for the  backward doubly stochastic  system with time delay under certain conditions.

Using Jordan standard form of generalized cubic matrix, we gave the necessary and sufficient conditions that the generalized cubic matrix solutions to the matrix equation AX=A+X existed and the form of the soluti
ons, and proved that the solutions to the absolute value equation Bx-|x|=b determined by the generalized cubic matrix solution B to AX=A+X existed.

According to the principle of epidemic dynamics, we established a class of SIRS epidemic model based on  psychology effect. The sufficient conditions for the global asymptotic stability of the model at the disease free equilibrium point and endemic disease equilibrium point  were obtained by constructing Lyapunov function, using the stability of ordinary differential equation and the theory of limit system, and then the results 
were verified  by numerical simulation. The  results of numerical simulation show that the positive influence of psychology effects can control the spread and prevalence of infectious diseases to a certain extent.

We used least square QRfactorization (LSQR) method to study the problem of structural damage identification. Firstly, we proposed a damage location method. Secondly, the damage identification problem based on the generalized modal flexibility matrix was transformed into a least squares problem. Finally, the LSQR method was used to progressively determine the location and extent of damage. The effectiveness of the proposed method was verified by two examples.

We  define and study a class of submodule of Z_pZ_ps［ξ］, that is the skew constacyclic code over  Z_pZ_ps［ξ］, where   Z_ps［ξ］ and   Z_p［ξ］are Galois rings. Firstly, we discuss the generator matrix of   Z_pZ_ps［ξ］ linear code. Secondly,  we give the generating polynomial set of skew constacyclic code over   Z_pZ_ps［ξ］ by discussing the generating polynomial of skew constacyclic code over   Z_ps［ξ］. Finally, we give one example.

Let m,n be two given positive integers and R be a ring. By introducing the notion of (m,n)-pure hereditary rings, we gave some equivalent characterizations of (m,n)-pure hereditary rings by using the homological method.

Let M be a factor von Neumann algebra on a Hilbert space H in which the dimension of M is larger than one. We give the definition of nonlinear *-Lie triple derivation on M. By the method of Pierce decomposition, it is proved that if Φ: M→M is a nonlinear *-Lie triple derivation, Φ is a nonlinear *-Lie derivation.

We extended the quadratic Novikov superalgebras to quadratic HomNovikov superalgebras by a twisting map. 
When the twisting map in HomNovikov superalgebra was an automorphism or an involution, we gave the relationships between quadratic Hom-Novikov superalgebras and quadratic Novikov superalgebras, established the relations between quadratic Hom-Novikov superalgebras and quadratic Hom Lie superalgebras, proved that quadratic Hom-Novikov superalgebras were Hom associative algebras, and the adjacent Hom Lie superalgebras of quadratic Hom-Novikov superalgebras were 2step nilpotent.

By using the method of ordinal sum, we investigated the  necessary and sufficient conditions for a general fuzzy implication to be a proper complement to linear transformations of a given family of fuzzy implications on a given family of lower triangular parts of subsquares in the unit square. We put all kinds of  lower triangular ordinal sum implications into a unified framework, and then gave the general form of construction of  lower triangular ordinal sums  of fuzzy implications.

Firstly, by using the method of conditional least squares, we discussed the problem of parameter estimation for MGINAR(p) model with missing data, and obtained the conditional least square estimations of parameters. Secondly, numerical simulation was used to verify the feasibility of four methods for processing missing data and to compare the estimation effects. The simulation results show that when the missing probability is small, the case rejection method or the mean value interpolation method can be used, and when the missing probability is large, the bridge interpolation method can be used to reduce the estimation bias.

We constructed a hostparasite epidemic model with vertical transmission. Firstly, the local stability and global stability of model were analyzed by Jacobian matrix and BendixsonDulac theory. Secondly, the basic reproduction number of the model was given. Finally, the results were verified by numerical simulations. 
The results show that the vertical transmission of parasites can reduce the density of the host, but will not lead to   extinction of the host population.

The square G² of a graph G is a graph such that V(G)=V(G²) and uv∈E(G²) if and only if the distance between u and v is at most two. The chromatic number χ(G²) of G² is the minimum k for which G² has a proper k vertexcoloring. Using the discharging method, the author proves that if mad(G)<4 and Δ(G)≥7, then χ(G²)≤3Δ(G)+1,  if mad(G)≤4 and Δ(G)≥8, then χ(G²)≤3Δ(G)+5.

By using the construction of coloring method and greedy algorithm, we studied the b-coloring number and m-degree of corona product and edge corona product of any two graphs, two graphs of the same order and any two complete graphs for limit maximum and minimum degree. On the basis, we derived the b-coloring number and mdegree of the generalized corona product and edge corona product for partial graph class.

Let a,b be two generalized Drazin invertible elements in a Banach algebra, a^d be the generalized Drazin inverse of aaⁿd a^π=1-aa^d. We  gave the expressions for the generalized Drazin [JP2]inverse of the sum a+b  under some conditions such as ab^π=a, b^πba^π=b^πb, b^πa^πa²b=b^πa^πaba, b^πa^πb²a=b^πa^πbab etc by using the idempotent element in a Banach algebra.

Aiming at the problem that the traditional similarity calculation method only used the user’s rating information as the explicit feedback behavior to recommend, which led to the unsatisfactory recommendation effect, we proposed a new similarity calculation method to improve the accuracy of the recommendation by introducing implicit feedback information, such as weight adjustment mechanism and user behavior preference. Firstly, according to the anti-user frequency of negative sampling, the influence degree of global software engineering of popular items was reduced, and the minimum weight of the common scoring behavior was used to adjust the recommended accuracy deviation caused by few common scoring. Secondly, we proposed the definition of project preference words, and calculated users with common preference in project characteristics according to the matrix of project preference words. Finally, the experimental comparison and analysis on the MovieLens data set. The experimental results show that the improved similarity calculation has better MAE value and higher recommendation accuracy.

In view of the problems of traditional facial expression recognition algorithms, such as poor feature extraction ability, low recognition rate and high misclassification rate, we proposed an improved facial expression recognition algorithm based on generative adversarial network (GAN). Using the game theory of generating antagonism network, the feature extractor, feature combiner and discriminator were designed respectively. Through the adversarial training between the discriminator and feature extractor, the feature extraction ability of the feature extractor and the accuracy of the classifier for facial expression recognition were continuously enhanced, and it was applied to the intelligent monitoring of working state of staff. The working state was judged according to the results of expression recognition, so as to reasonably allocate the laboratory resources and improve the utilization rate of laboratory resources. The several experimental results of the improved algorithm on the CK+ dataset show that the proposed algorithm has higher robustness and can effectively improve the rate of facial expression recognition.

Aiming at the problem that it was difficult for the traditional algorithms to meet the requirements of accuracy and timeliness of ground station resource scheduling, we proposed a hybrid decomposition algorithm which combined the heuristic algorithm based on conflict degree evaluation index of ground station resource usage and Lagrangean decomposition algorithm to solve the problem, so as to speed up the search of the optimal solution. The results of simulation tests show that the proposed algorithm can get high quality solutions to the resource scheduling problems of remote sensing satellite ground stations in a short time.

In order to improve the adaptability and control accuracy of k-anonymity model for a-diversity large data delay scheduling, we proposed an adaptive delay scheduling algorithm for a-diversity k-anonymity large data based on wavelet analysis. Firstly, wavelet analysis was used to denoise the data, and priority list control model was constructed by denoising data. Secondly, a channel model of load balancing transmission was designed by using efficient time division multiple access protocol, and an objective function of scheduling control was established by combining adaptive weighting control method. The optimal solution of the objective function was achieved by slot allocation to realize the improved design of scheduling algorithm. The simulation results show that the adaptive equilibrium performance of a-diversity k-anonymity large data scheduling using the proposed method is better, the relative error of data scheduling is low, and the recall rate of data is better than the traditional method.

Aiming at the current problem of single basic color features of image and insufficient information available, we proposed to transform the image into HSV color space. We first used the equalization histogram to enhance the spatial color pixels of the image after the spatial color conversion, and then used K-means clustering algorithm integrated with rgb2ind to brighten the balanced image, so as to enhance the extraction target and quantity of image color feature. The experimental results show that the proposed method is better than that of the ordinary spatial color image for feature extraction, and provides more retrieval information for color feature recognition.

Aiming at the problem of large amount of work on the value of clustering number K in partition clustering algorithm, we proposed a new Canopy+ algorithm. The proposed algorithm can predict the clustering number K and improve the clustering efficiency on the premise of ensuring the accuracy.

Aiming at the problems that current image segmentation methods were difficult to achieve accurate and fast image segmentation, we proposed a fast image segmentation method for controlling the evolution of active contour. Firstly, the weighted sum of external energy and internal energy was used as the curve energy function, and the active contour wave model was established by using the external and internal energy
 of closed curve. Secondly, the curve energy was updated by moving the optimal path to obtain the required image segmentation target. Finally, the particle swarm optimization algorithm was introduced to obtain the optimal control points of all the initial contour points, and the goal of accurate segmentation of the target image was achieved by controlling the evolution of the active contour according to the optimal control points. The experimental results show that the image segmentation accuracy of the method is significantly higher than that of the current typical image segmentation method, which improves the anti noise performance and the speed of image segmentation.

Aiming at the problem that the scattered data was easy to lose frame when the data was randomly forwarded in the integrated data interaction center, the author proposed a lossless restoration method based on big data analysis. Firstly, big data bit sequence was formed by collecting scattered defect information in the data interaction center of cloud storage, and the storage structure model of data in cloud computing center was analyzed. Secondly, the feature decomposition of scattered defect information was carried out by using the method of joint sparse decompo
sition,  the feature compression and adaptive feature clustering processing of information of scattered defect information were carried out by using compression perception method, combined with big data information fusion method, and the association feature mining of scattered defect information was carried out. Finally, the method of phase space reconstruction was used to reconstruct the features of scattered defect information, and the lossless information of scattered defect information was restored in the reconstructed phase space. The simulation results
 show that the error of lossless restoration of scattered information is lower, the precision of data reconstruction is higher, the computation cost is small, and the ability of information restoration is improved effectively.

Aiming at the problem that the traditional neural network proportionalintegraldifferential (PID) controller and linear quadratic regulator (LQR) optimized PID controller had long recovery time and poor antiinterference for speed control of brushless direct current motor， we proposed a BP neural network PID controller optimized by LQR for speed control of brushless direct current motor. Firstly, BP neural network was used to adjust the PID gain to improve the dynamic adaptability and robustness of the controller. Secondly, LQR was used to optimize the optimal output of BP neural network, which was closer to the target PID gain. The simulation results show that the controller can effectively improve the response speed, reduce the steadystate error and enhance the antiinterference ability.

We proposed an intelligent hybrid controller based on traditional PI and adaptive fuzzy PID (proportion integration differentiation) to improve the speed tracking and dynamic adaptability of brushless direct current motor. The controller switched flexibly between the steadystate PI controller and the transient adaptive fuzzy PID controller according to the speed error through the switching function. The adaptive fuzzy PID controller included two levels of fuzzy logic controllers, the firstlevel fuzzy logic controller adaptively adjusted the PID gain, and the secondlevel fuzzy logic controller automatically adjusted the scale factor of the firstlevel fuzzy logic controller. The simulation tests were carried out in
the MATLAB/Simulink environment. The test results show that compared with traditional PID and fuzzy PID controllers, the speed response, steadystate error and overshoot/undershoot of the intelligent hybrid controller under various conditions are greatly improved,
 indicating that it has better robustness, antiinterference and dynamic adaptability.

Aiming at the problems of inaccurate retrieval results and low efficiency of the traditional self-integrated information retrieval method in the retrieval of the selfintegrated information of books, we proposed a fast retrieval method for selfintegrated information of books based on Top-k query algorithm. Firstly, the Top-k query algorithm was used to match the denoising selfintegrated information. Secondly, a fast retrieval model of selfintegrated information was established to quickly retrieve the selfintegrated information of books according to the importance of word meaning. The simulation results show that the proposed method has good retrieval performance, the overall operation time is shortened, the retrieval efficiency is improved, and the retrieval results are more accurate.

We considered a fourbody coupled cavity array with symmetric gain and dissipation, and analyzed the PT phase transition and the single photon transmission of the system. The results show that the system can be in thePT symmetric phase and the PTsymmetric broken phase according to different parameter ranges, which the photons are located in different cavities, and the single photon transmission is bidirectional.

We calculated the mass of the neutron star by using seven parameter sets and considering the hyperon degrees of freedom in the framework of the relativistic mean field theory (RMFT), and analyzed the influence of coupling constants of hyperons on the maximum mass of neutron star. The results show that adopting three type sets （NLSH,NL3,NL2）in accordance with relative stiff equation of state (EOS), the maximum mass of neutron star can reach the mass given in the observed massive neutron star PSR J0348+0432,PSR J16142230 and the binary neutron star merger GW170817, the massive neutron star can contain hyperon phase. With the increase of the hyperon coupling constant (χ), the equation of state stiffens, and the maximum mass of neutron star increases. When χ≥0.6 of the NLSH parameter set, the maximum mass of neutron star is more than 2M⊙（M⊙ is mass of the sun）. Therefore, the massive neutron star can be obtained by choosing relative stiff EOS from three parameter sets and relative strong coupling constants of hyperons.

Using the B3LYP method of density functional theory, the author optimized ground state stable geometries of threonine (Thr) molecules and charged ions at the level of 6-311+G(2-df) basis set. The method of time dependent density functional theory  (TD-DFT) was used to study the excited states of the Thr molecule and ion system under the implicit solvent ethanol. The results show that with the increase of the number of electrons captured by molecules, the energy band of the molecule system decreases slightly. The S1—S10 excited states of Thr2- ion system are all obtained from the occupied orbital of MO33. The excitation characteristics of Rydberg are good, and the excitation mode of the excited stste of ion system which captures electrons changes significantly.

The first principle was used to calculate the energy band and electronic density of states of N and S atom codoping, and to study the
 relationship between quantum capacitance and different doping configuration. The results show that the introduction of N/S and N/B codoping atoms can lead to localized state at Fermi level. In the range of -0.6 V to 0.6 V potential, the quantum capacitance of NSS,NS,NBB,NNB and NB doping structures is higher than that of pristine silicene. The quantum capacitance of NSS doping single vacancy silicene is 43.9 μF/cm² at Fermi level.

We prepared ZIF-67 metal organic framework with copper partially replacing cobalt, and introduced Pt nanoparticles into ZIF-67 pore c
hannels. The samples were analyzed by scanning electron microscopy (SEM), transmission electron microscopy (TEM) and Xray diffraction (XRD), and the electrochemical sensing performance was tested. The results show that the electrode modified by the noble metal and copper comodified ZIF-67 metalorganic framework has good crystallinity and uniform size. The enzymefree electrochemical sensors has good detection performance for hydrogen peroxide, such as high sensitivity, wide detection range, excellent selectivity and good repeatability.

Y2O3∶Er3+ ，Yb3+ nanocrystals were synthesized at different
pH values and various urea concentration by homogeneous coprecipitation method after 700 ℃ sintering. The phase structure, microstructure and luminescent properties of the samples were characterized by Xray diffraction (XRD), transmission electron microscopy (TEM), Fourier transform infrared spectra (FTIR) and fluorescence spectrometer. The results show that the pH value has great effect on the particle size of Y₂O₃∶Er³⁺,Yb³⁺ and the particle size increases with the increase of pH value. The upconversion fluorescence intensity of green (500—600 nm) and red (650—700 nm) is obviously enhanced, and redgreen ratio is gradually reduced. The effect of urea concentration on Y₂O₃∶Er³⁺,Yb³⁺ nanoparticles is small.

We used TDNA insertion method to construct a mutant library of Setosphaeria turcica, and screened a mutant with significantly decreased pathogenicity. T-DNA insertion site of the mutant was upstream promoter region of the nonribosomal peptide synthetases 6 (StNPS6) gene. We first knocked out the StNPS6 gene, and then analyzed the differential expression analysis of transcriptome and proteome of StNPS6 gene knockout mutants and wild types. The results show that  the 903 genes are upregulated and 864 genes are downregulated
among 1 767 differentially expressed genes. The 8 proteins are upregulated and 22 proteins are downregulated among 30 differentially expressed proteins identified in the mutant.

Using prokaryotic system to express bovine viral diarrhea virus (BVDV) E2 protein, we prepared mouse polyclonal antibody. RNA was extracted by MDBK cell proliferation virus, the fulllength E2 gene was amplified by RTPCR, truncated primers were designed to construct the optimal expression vector pET-30-E2 and transformed BL21 after bioinformatics analysis. The protein expression was induced by IPTG, the protein expression was analyzed by SDSPAGE electrophoresis, and mice were immunized with purified protein and Freund’s adjuvant. Enzyme linked immunosorbent assay (ELISA) was used to determinate the antibody titer level, western blotting (WB) and immuno fluorescence assay (IFA) were used to ver
ify the specificity of the antibody. The results show that E2 gene was successfully expressed in E.coli, the protein size is 32 000 and it
is insoluble. The protein content is 0.4 mg/mL, the titer of polyclonal antibody is 1∶256 000, and it can specifically bind to the recombin
ant protein and virus in cells.

Ultrasonic extractionhigh performance liquid chromatography was adopted to determinate the mass concentration of nitrification inhibitor—nitrapyrin in different soil types using the Agilent SBC18 (150 mm×4.6 mm, 5.0 μm) chromatographic column under the condition of optimizing the best volume ratio and the flow rate of the mobile phase. The results show that the best extracting effect is that the soil sample is disposed by the organic solution of ethanol for three times continuously followed by organic filter membrane to purify, then chromatographic column of Agilent SBC18 is separated under the condition of the mobile phase being V(acetonitrile)∶V(water)=4, and completing within 10 min at the flow rate of 0.8 mL/min. There is a significant linear relationship between the mass concentration of nitrapyrin and the peak area in the range of 0—75 mg/L (R2=0.999 8, p<0.01). The detection limit of nitrapyrin in blank soil is 0.42 mg/L. The recovery rate of nitrapyrin for different mass concentrations and dosage forms in meadow black soil, chernoze and aeolian soil is 82.44%--91.66% after 24 h of addition with the standard deviation of 1.34—5.04 and the coefficient variation of 0.02—0.06.

Yttrium ion (Y³⁺)doped titanium dioxide (TiO₂) pillared bentonite was prepared by the acid sol method, and it was used to photocatalytic degradation of methyl orange (MO). The intermediate products and catalysts were analyzed by ultravioletVisible spectrophotometry (UV-Vis), liquid chromatographymass spectrometry (LC-MS), and ultravioletVisible diffuse reflectance spectroscopy (UVVis DRS) during the degradation process, the changes of chemical oxygen demand (COD), total organic carbon (TOC), conductivity, and pH value during degradation were studied. The results show that after 6 h of reaction, the solution is basically colorless, and degradation rate is 99.6%. During the reaction, toluenesulfonic acid, aniline, hydroquinone and other intermediate products are generated, which continue to degrade and mineralize as the reaction proceed. To the end of the reaction, the pH value of the solution decreases from 5.0 to 4.6, the conductivity increases from 82.60 μS/cm to 14150 μS/cm, the COD value of the supernatant decreases from 80.20 mg/L to 12.20 mg/L, [JP+1]the COD value of the suspension decreases from 201.20 mg/L to 102.90 mg/L, the TOC value of the solution decreases from 30.20 mg/L to 13.10 mg/L, and the TOC value of the catalyst decreases from 76.80 mg/L to 43.10 mg/L. The reaction process is carried out according to the synergistic mechanism of adsorption photocatalysis.