By using the accurate estimates of the  Hormander class, we proved the boundedness of the commutators generated by the bilinear pseudodifferential operators and Lipschitz functions and BMO functions on the generalized Morrey spaces, and then obtained the boundedness of the commutators of the bilinear pseudodifferential operators on the classical Morrey spaces.

By using a bootstrap argument method, we considered the global existence for the solution of a class of fractional Hartree equations with timedependent damping/gain. We discussed the global existence condition of the solution from two aspects of the damping/gain coefficients and the initial value, respectively. We obtained a global existence condition of the fractional Hartree equation which only depends on the damping/gain coefficients and only depends on the initial value of the solution.

We considered the Caputo fractional differential equation with RiemannStieltjes integral boundary condition. We established the existence theorems of positive solutions for the integral boundary value problem of fractional differential equations under the condition of allowing the singularity of the nonlinear term, and proved the existence theorem by using mixed monotone operator method and fixed point theorem on semiordered set. Some examples show the applicability of the conclusion.

Based on the MoralesRamis theory, we considered the nonintegrability of the Hamiltonian system with Marta potential energy by using the theoretical analysis method, and proved that this Hamiltonian system was not meromorphic integrable in sense of Liouville. The results can be used to understand the dynamic behavior and topological structure of the system from the perspective of nonintegrability.

By using mountain pass lemma and Lion’s lemma, and combined with Pohozaev identity, we got the existence of nontrivial ground state solution for the nonlinear fractional Schrdinger system.

We considered the properties of  blowup of solutions of energy for a class of viscoelastic wave equations with variable\|exponents. By constructing an energy function, we studied the properties of the energy function, and used the obtained  results, Cauchy inequality and integral estimates to get the   properties of blowup of solutions of energy for a nonlinear wave equations with variableexponents in finite time.

Using the methods of boundary function and fractional steps, the author constructed the asymptotic solution of steptype  contrast structure for a class of weak nonlinear singularly perturbed differentialdifference equation, and obtained the time t=t* to produce a contrast structure. By means of sewing method, the author proved the existence of the smooth solution of interior layer and the uniform validity of the asymptotic solution of the form.

We considered the spatial dynamic behavior of a predatorprey system with time delay and nonlinear prey harvesting effect. First, we obtained the condition of Hopf bifurcation and Turing bifurcation by using stability theory and bifurcation theory, and then numerical simulations show that the system has rich dynamic behavior. Numerical results show that time delay and diffusion can not only affect the formation of Turing patterns, such as spot, stripe and the coexistence of the two types, but also affect the formation of spiral patterns.

By using a new algebraic inequality, we established two inequalities for generalized normalized Casorati curvatures of submanifolds in a Bochner Kaehler manifold,which gave relationships between the normalized scalar curvatures and the extrinsic invariants Casorati curvatures of submanifolds.

We considered A-Browder’s theorem and A-Weyl’s theorem of Weyl theorem. By using the topological uniform descent method, we obtained a new equivalent condition and decision method for arbitrary C∈B(H), operator M_C satisfying A-Browder’s theorem and the A-Weyl’s theorem under small compact perturbations.

We considered some properties of maximal annihilators of WGPinjective rings, and proved that: if a left WGPinjective ring R satisfied that the ascending chain l(a₁)l(a₁a₂)l(a₁a₂a₃)… consisting left annihilator were stable for arbitrary infinite element a₁,a₂,a₃,…,  then R was a semiprimary ring; R was a left and right perfect ring; and R was a left and right Kasch ring. Some sufficient and necessary conditions of WGPinjective rings were given.

A particular solution method was used to compute a class of time harmonic medium scattering problems. Based on a simple partition of the domain, fundamental solutions were selected to approximate the property of the scattered field at local points and infinity. There was no need to cut unbounded domain into a bounded domain. The algorithm was suitable  for the general scatterers and multiple scatterers, which could achieve high precision in rough grids and achieve fast convergence.  Numerical examples show  the feasibility  and effectiveness of the algorithm.

We proposed a new labelling called the generalized edge magic graceful totally labelling, and used a few algorithm of generalized edge magic graceful labelling to encrypt and optimize the Apollonian network models. We chose the algorithm that made the number of magic constants be minimal estimate, and obtained the upper bound.

Let {X_n, n≥1} be a strictly stationary sequence of ρ-mixing random variables. Using the central limit theorem and moment inequality of ρ-mixing random sequences, under suitable conditions, we gave general function formula for the precise asymptotics of  the complete moment convergence of the  partial sums  for ρ-mixing random variable sequences in the logarithm law.

The author considered  the existence of positive solutions for the mixed boundary value problems of the onedimensional p-Laplacian equations  with nonlinearities changing signs, by using the nowhere normaloutward compact map’s fixed point index theorem, under the condition of f(t,0)≥0 and p-sublinear conditions, the author obtained the existence of the positive solution of the problem.

We invesigated the oscillation of the solution of the secondorder mixed partial difference equation with two constant coefficients pU_m+2,n+qU_m,n+2-U_m,n+U_m+σ,n-τ=0 by using the envelope theory, and obtained three necessary and sufficient conditions for the oscillation of solutions of the equation,  where σ and τ are positive integers, m and n are nonnegative integers, p and q are real numbers.

Based on the invariant manifolds theory, we obtained the relationship of integrability between the high dimensional Kolmogorov systems and its subsystems by  analysing the retention of the first integral, and gave the decision results of integrability of a class of twodimensional Kolmogorov systems.

We used the measure theory and truncation method to disscuss the properties of solutions for a class of degenerate elliptic equations with weighted functions. The result shows that the approximated solutions of the problem tend to zero uniformly.

The finite state automata was used to model distributed discrete event systems, then we proposed an  algorithm of safe diagnosability for distributed discrete event systems. First, we got the global model of the system, pruning and determining whether the given system could be diagnosed or not. If the system could be diagnosed,  the safediagnostor of the system was constructed. Then the safe labels were added to the global model to determine whether the system was safe diagnosable or not. The experimental results show that the algorithm can judge whether the distributed discrete event system is safe diagnosable or not, and the solution space is reduced. The state numbers are approximately reduced by 7 times in the best case, and the state numbers are approximately reduced by 5.45 times in the average case.

In order to quickly and accurately partition the community of large\|scale social networks which were increasingly complicated, we proposed an improved algorithm based on similarity voting to replace the underlying partition of Louvain algorithm. It solved the shortcomings of Louvain algorithm such as slow convergence in the bottom partitioning and large number of double counting, which made the community partition more rapidly. The experimental results from real social network data show that compared with the Louvain algorithm, the efficiency of the improved algorithm is much higher, with less number of communities partitioned, and the community structure is more compact in the case of keeping the modularity basically unchanged.

We proposed a multitexture centrosymmetric local binary pattern (CSLBP) feature to realize multiview face detection in complex environments. The feature retained the characteristics of Haar ordinal relations, so we drew on the experience of the combination of local binary pattern (LBP) to extract features from four texture directions, such as horizontal, vertical, +45° and -45°, so as to ensure the robustness of face detection in direction, illumination, rotation and so on. The algorithm adopted the cascade architecture. First, face images were partitioned according to different angles of view, and multitexture features were extracted respectively. Then some independent classifiers were designed to eliminate the nonface window step by step. Finally, the multilayer perceptron (MLP) was
used to synthesize the detection effect of each angle of view to output the detection results. The results of verification experiments on data sets FDDB and CMU PIE show that this method is effective for multiview face detection in complex environment. Compared with traditional convolution neural network face detection method, this method has higher accuracy.

We proposed a remote sensing images fusion algorithm based on weighted wavelet analysis. Firstly, we extracted the illumination intensity component of multispectral images in the illumination chroma saturation space, and made a principal component analysis of the illumination intensity component to get the corrected illumination intensity. Secondly, the weighted fusion was carried out by corrected illumination intensity and wavelet analysis for the high frequency region. Finally, the final fusion results of remote sensing images were obtained by inverse transform of wavelet analysis and illumination chroma saturation space. This algorithm effectively solved the problems of block blurred phenomenon caused by the way of wavelet analysis to discard low frequency components and distortion of information domain produced by principal component analysis in the process of image fusion with low spatial resolution and high spatial resolution. We carried out simulation experiments on the fusion of remote sensing image in different scenes. The results show that the proposed algorithm can solve some common problems such as blurring of edge after image fusion and fuzzy block shadow appearing in fusion result. It is greatly improved in clarity, texture details and authenticity, and can adjust the appropriate weighting coefficient for different fusion needs, so that the fusion of remote sensing images can achieve the best effect. The weighted wavelet analysis further improves the effect of remote sensing image fusion, not only fully expresses the details of various remote sensing images, but also preserves the orginal spectral information better.

We proposed a hybrid kernel support vector machine (SVM) algorithm based on improved particle swarm  optimization (PSO) algorithm, which solved the problem that the general hybrid kernel SVM algorithm was difficult to evaluate the parameter selection. The algorithm improved the convergence property by limiting the velocity of the particle, the search space and the crossover operator to get the best combination of the parameters. Simulation experiments show that the algorithm can get the optimal value of parameters more quickly and effectively.

We used a distributed fuzzy controller method to carry out the fault tolerance and nonuniform clustering for wireless sensor networks. First, the residual energy of nodes, centrality of nodes and distance from nodes to base station were input to the fuzzy controller. The IFTHEN rule was used to reason and output “cluster head opportunity” and “cluster size” value to make the best performance node become cluster head and build a suitable cluster size. And then data transmission by time division multiplexing mode could tolerate the temporary and permanent faults of the cluster heads and member nodes. The simulation results show that the algorithm can effectively balance the load among clusters, reduce the energy consumption of the network, and thus improve life cycle of the network.

Aiming at the problems that the matching threshold was difficult and the recognition speed was slow in the process of vehiclelogo recognition, we proposed a vehicle[KG-*4]\|logo recognition algorithm based on feature matching of scale invariant feature transformation (SIFT). The SIFT operator was used to extract the invariant features of the image, such as viewing angle, translation, radiation, brightness and rotation, and the BP neural network algorithm was used to autonomously select vehiclelogo image features for classification, matching and recognition. The results of simulation experiment show that the mean values of recognition rate for simple vehiclelogos and complex vehiclelogos are all more than 90%, the algorithm  has faster recognition speed and higher recognition rate, which can meet the needs of practical application.

In order to realize the adaptive software, we proposed an adaptive software decision model based on hidden Markov model (HMM) for the complex and changeable operating environment.  Firstly, the Gaussian mixture model (GMM) was used to classify the initial environment. Secondly, the softmax regression was used to classify and divide the perceptual environment. Finally, we used HMM instead of the manual intervention to make software decisions. The experimental results show that the adaptive software model can achieve the adaptive software decision well under the condition of the change of the perceptual environment.

In order to obtain ideal result of node localization, we designed a wireless sensor node localization algorithm based on ranging distance modified by particle swarm. First, the working principle of the classical wireless sensor node localization algorithm (DVHop) was analyzed to find the factors that lead to the ranging distance error, and then, particle swarm optimization algorithm was used to modify the ranging distance between the nodes of wireless sensor to reduce ranging distance error among the nodes in which standard particle swarm optimization algorithm was improved correspondingly. Finally, the simulation experiments were compared with the classical node localization algorithms of wireless sensor. The test results show that, in the same working environment, the proposed algorithm improves the location precision of wireless sensor nodes, and does not increase the extra hardware overhead.

By using certificateless public key cryptosystem, we constructed a new aggregated signature algorithm to achieve the authentication of information transmission between vehicle nodes in vehicular sensor network. The new scheme achieved the conditional privacy protection for vehicle users through shadow value and nodes around the roadside units generated. The experimental results show that the scheme not only eliminates the complexity of certificate maintenance costs, but also solves the key escrow problem, and meets the authentication, anonymity, unforgeability and traceability of message, as well as the communication efficiency is higher, and the message verification delay is shorter.

Aiming at the information security problems of invehicle network, on the basis of anomaly detection method of the controller area network (CAN) bus, we
proposed an anomaly detection method for CAN bus message based on the random forest model. Firstly, a large number of normal and abnormal message data were used to construct a random forest model and perform a series of parameter adjustments. Secondly, the CAN bus message to be detected was input into a random forest model of the corresponding ID. Finally, a classification of the normal or abnormal message was completed by the model.  The results of simulation experiment show that the model can effectively detect the abnormal data on the bus, and improve the safety of the vehicle operation.

Aiming at the problem of low accuracy of 3D model feature recognition in massive data mining, we proposed an improved sparse denoising autoencoder neural network model. Firstly, based on the improved sparse denoising autoencoder method, we constructed a deep neural network model, then used the unsupervisedpretraining method and restricted quasi Newton method to train the autoencoder neural network. Finally, the softmax regression and the obtained features were used to train the final classifier. The results show that the method has well robustness to the feature information of the three\|dimensional model with noise. Compared with the stack of autoencoder neural network and the selflearning neural network, the method has better recognition rate.

In view of brain tumor images, we proposed a retrieval method based on content. Firstly, we augmented the region of interest (ROI) of brain tumor images by using the image expansion algorithm. Secondly, based on the feature extraction method of pointbased spatial pixel intensity, we described the local characteristics of the augmented ROI. Finally, we introduced the aggregated feature algorithm to improve the discrimination and compression ability of image features. The results of the experimental comparison of three brain tumor image retrieval methods show that the algorithm can obtain high retrieval accuracy and prove that the algorithm can achieve efficient and accurate brain tumor image retrieval.

Aiming at the problems of slow convergence rate and low positioning accuracy of the indoor acoustic source localization algorithms under the condition of strong noise and strong reverberation, we proposed an acoustic source localization method based on improved time delay estimation. The method was based on a single source multiple reverberation model. Firstly, the quaternary cross microphone array was used to estimate the time delay. Secondly, on the basis of the generalized crosscorrelation time delay estimation algorithm, quadratic correlation method was introduced to weaken the noise interference. At the same time, the least mean square (LMS) adaptive filtering algorithm was used to compensate for the deficiencies of the generalized crosscorrelation method and improve the time delay estimation precision in the reverberation environment. Finally, the acoustic source was located by the farfield approximation geometric method. Experimental results show that,  compared with the generalized crosscorrelationphase transform (GCCPHAT) algorithm, the proposed method has better antinoise and antireverberation abilities, and achieves more accurate positioning results.

Based on the principle of XML configuration, we gave a greedy strategy algorithm and tested the algorithm. Aiming at the problems existing in the testing process, we optimized the algorithm to obtain the multidimensional extended greedy algorithm, and carried out comparative experiments on different algorithms. The experimental results show that the algorithm has the advantages of high reliability, fast execution and comprehensive coverage, and the more data are run, the more accurate the results are.

In order to obtain the ideal hole repairing result of point cloud data, aiming at the defects existing in the current algorithms, we proposed a hole repairing algorithm for point cloud data based on least square support vector machine. First, the hole repairing range was estimated according to the boundary of scattered point cloud, and then according to information of hole and surrounding points, we built a surface by least square support vector machine, and repaired the hole in the point cloud data. Finally, the simulation experiment was realized by C++ language programming. The experimental results show that the least square support vector machine can effectively repair various complex holes, and the repair effect is better than other algorithms.

We investigated the twodimensional spatiotemporal dynamical behavior evolution and spatiotemporal chaos of the photorefractive ring oscillator under the condition of asymmetric couple by using the uniform external field modulation to control spatiotemporal chaos of asymmetric couple photorefractive ring oscillator. The results show that the photorefractive ring oscillator can be controlled into periodic 4, periodic 2 and periodic 1 with the increase of modulation intensity through inverse bifurcation. The chaotic laser output of ring cavity laser is used as modulation signal, and the photorefractive ring oscillator can be controlled into periodic 4 and periodic 2 by adjusting the gain coefficient of the ring cavity laser.

We studied the classical entanglement problem of four parity-time (PT) symmetric Bell states under Hermitian conditions. The results show that among
the four PT symmetric Bell states, |Φ_-〉 and |Ψ_-〉 are still classical quantum states, and are still maximally entangled states under Hermitian inner product; whereas |Φ₊〉 and |Ψ₊〉 are no longer always classical quantum states, and the classical quantum states normalized by Hermitian inner product are no longer always maximally entangled states under Hermitian inner product.

Based on sliding mode control and proportional integral sliding mode control, the author designed sliding mode functions and controller, and gave sufficient conditions for synchronization of fractionalorder NewtonLeipnik chaotic systems. The results show that the masterslave systems of fractionalorder Newton Leipnik systems obtain sliding mode synchronization and proportional integral sliding mode synchronization if proper control law and sliding mode surfaces are selected.

Using the B3P86 method of density functional and timedependent density functional theory (TD-DFT), we calculated the excitation energies, wavelengths and oscillator strengths of acetonum molecule from the ground state to the excited states of 1—10 at the level of 6-311++G(2-df-) basis set, and investigated the effect of the external electric field on the  excitation state acetonum molecule. The results show that the excitation energy  is greatly influenced by the electric field intensity. The acetonum molecule can be easily excited and dissociated under the action of external electric field.

Based on the density functional theory, we investigated the chiral transition process of the asparagine molecule in 1F-MOL at the level of ONIOM (CAM-B3LYP/6-31G(-d,p-): UFF) basis set. We drew a complete chiral transition path reaction potential energy surface by finding the structures of the extreme value points including the transition states and intermediates, and analyzed the geometric and electronic structure properties of extreme value points. The results show that the hy
drogen atom on the chiral carbon atom of Sasparagine molecule can transfer to the other side of the carbon atom via the oxygen atoms of carboxyl group as
a bridge, and achieve the chiral transition of asparagine molecule from Stype to Rtype. The maximum reaction energy barrier is 321361 2 kJ/mol, which comes  from the second transition state TS₁-R-Asn@1FMOL.

We investigated the chiral transition mechanism of the ibuprofen molecules confined in the helically chiral SWCNT by using the ONIOM methods of combination of quantum mechanics and molecular mechanics. The results show that the deformation of ibuprofen molecules in the confinement region is more and more obvious with the decrease of chirality SWCNT diameter. There is only one reaction channel for the optical isomerism of ibuprofen molecules in SWCNT(6,4) and SWCNT(7,4), and there are two reaction channels for the optical isomerism in SWCNT(8,5). When the ibuprofen molecules are confined in SWCNT(6,4), SWCNT(7,4) and SWCNT(8,5), the energy barriers of ratedetermining step for the optical isomerism reaction are 24795,27383,29224 kJ/mol, respectively, while the total energy barrier are 278.96,291.91,325.88 kJ/mol. The optical isomerism of S-Ibu molecule is easy to be achieved in the smaller pore size of the helically chiral SWCNT, and SWCNT(6,4) can be used as a nanoreactor for optical isomerism of ibuprofen molecules.

The natural polymer oligochitosan was modified using o-vanillin. By the excellent coordination properties of the Schiff base structure, the copper complex of o-vanillic aldehyde modified oligochitosan was prepared. Cyclic voltammetry curves indicated that Schiff basecopper complexes possessed electrochemical activity, and the reaction of central Cu²⁺ at glassy carbonelectrodes was controlled by diffusion process mechanism in the range of scan rate from 0025—
10 V/s. We investigated the mechanism of the interaction of Schiff base and Cu complex with DNA by using cyclic voltammetry, ultraviolet spectrum, viscosity and melting point method. The results show that, after adding DNA, the oxidation peak current of the complex decreases and the peak potential moves weakly. VCOSCu13,VCOSCu31 and DNA can form complexes of molar ratio of 1∶1, and the binding constants are 0.78×104,0.69×104 L/mol. The intensities of absorption peaks of complexes are reduced and the peak position appears red shift. The relative viscosity and melting point of DNA increase after adding the complex. There are interca lation action between different molar of complex and DNA.

Based on the model of organism Caenorhabditis elegans, we investigated mass concentration of Phyllanthus emblica L.extracts at 10,20,40 mg/mL on antioxidation of C.elegans. The experimental results show that 10 mg/mL P.emblica L. extract can significantly increase the spawning of C.elegans, and there are no obvious effect on exercise capacity of C.elegans. The P.emblica L.extracts have low toxicity in the mass concentration range of experments. 10 mg/mL P.emblica L.extract can significantly prolong the life of C.elegans, and have a good protective effect on antioxidant pressure. 10 mg/mL P.emblica L.extract can significantly improve the enzyme activity of C.elegans, so as to achieve protective effect of antioxidant.

In order to improve the removal efficiency (interception rate) of boron of the reverse osmosis membrane in the treatment of the wastewater containing boron in nuclear power plants, by changing the pH value of the raw water and adding complex, we investigated the changing rule of the removal efficiency of boron of reverse osmosis membrane. The results show that the interception rate of boron in the wastewater containing boron by reverse osmosis membrane increases with the increase of the pH value. When pH=11, the interception rate is the highest. When n(sorbitol/mannitol)∶n(boric acid)=1, the interception rate of boron in solution by reverse osmosis membrane is the highest.

Taking the arseniccontaminated soil as the research object, the single and compound effects of 5 extractants were investigated by the batch oscillation of chemical extraction method. The optimum combination method of the best extractant in arseniccontaminated soil was obtained, and the ratedetermining steps and factors in the extraction process were investigated by the dynamic characteristic experiment. The results show that the removal effect of single extraction on arsenic in soil is oxalic acid(OX)>NaOH>ethylenediaminetetraacetic acid (EDTA)>KH2PO4>citric acid. The combination of EDTA and OX can achieve the highest arsenic removal rate with 855% under the optimized experiment conditions of m(soil)∶V(water)=1∶8, and oscillation time at 8 h. The appropriate twotwo combination of extractant for arseniccontaminated soil can achieve a better effect than single. There is no obvious difference in the effect of the combination of three and two extractants. The dynamic process of different compound extractants can be described by the quasisecondorder equation. It is speculated that the extraction process is mainly controlled by the chemical reaction mechanism, and the reaction involves electron sharing or electron transfer between the extractant and the arsenic compound.

The indoor culture method was adopted to conduct the aerobic decomposition of mixture composed of fallen leaves from Populus alba ‘Berolinensis’ L.& Hosta plantaginea Aschers.and turf grass clippings. Under different temperature and moisture content, we analyzed effects of two factors on the carbon mass ratio of water soluble substance, humicextracted acid, humic acid and humin (WSOC, w_C(HE), w_C(HA) and w_C(Hu) ), the E₄/E₆ of HA alkali solution and the w_C(HA) /w_C(FA) in the process of mixture decomposition. The results show that the molecular structure of HA tends to be simple at the end of culture, and the C(Hu)  can be effectively mineralized in the process of mixture decomposition; the w_C(HA)/w_C(FA) is the higher the temperature, the greater the decreased degree; the C(Hu) can be effectively mineralized and decomposed under different moisture content.