DNA-N6 methyladenine (6-mA) methylation modification is one of the most important epigenetic modification markers. The aberrant 6-mA modification can affect gene expression and lead to serious diseases. Therefore, the work of predicting the 6-mA site is of great significance for the understanding of the pathogenesis and treatment of diseases. In this paper, a long short-term memory (LSTM) neural network based on K-mer encoding method and one hot encoding method is proposed to predict methylation sites.Firstly, the information content of gene sequence is increased through K-mer coding method. Secondly, the information content after one hot encoding is converted into a composite encoding matrix. The LSTM model can extract more feature dimensions and types from the encoding matrix, to improve the prediction performance of the LSTM model for gene sequence. The cross validation experiment show that the proposed method can achieve accuracy of 93.7% on benchmark datasets. The sensitivity, specificity and matthews correlation coefficient of the trained model were 93.0%, 94.5% and 0.875, which outperformed existing 6-mA prediction methods. On the other six different species datasets, the proposed method can achieve the area under the curve (AUC) values from 0.9055 to 0.9262,which shows the applicability of the proposed method on animals, plants and microorganisms methylation tasks. The proposed method was applied on rice gene NC_ 029258.1, and the predictions were verified by the recently published online prediction tools. The results show that 86% to 96% of the prediction results are supported by these tools, indicating that the proposed method can be applied to large-scale site prediction and analysis of different species.

The classic recommendation system focuses on the accuracy of recommendation. With the increase of users’ diversified needs, the diversity of recommendation results has attracted more and more attention. The accuracy and diversity of the recommendation always conflict with each other, and the traditional recommendation algorithms often ignore the difference of users’ activity in the system. A user stratification multi-objective recommendation algorithm is proposed based on the difference of users’ evaluation times on the items, which can provide better recommendation results for different users. Users are divided into three types with high, medium and low evaluation times, and three different initialization methods are proposed for the algorithm. By the comparative analysis of the initialization operations and parameter settings of the existing probability based multi-objective evolutionary algorithm, improved crossover and mutation operations are obtained. The experimental results verify that the enhanced multi-objective evolutionary algorithm can find better results with higher accuracy and diversity. Finally, a recommendation scheme based on user stratification is summarized, which helps to improve the recommendation effect for different users.

To study the privacy protection of online car-hailing, an evolutionary game model including the government (regulatory, non-regulatory) and platform (self-discipline, non-self-discipline) is constructed, using the replication dynamic equation to find an evolutionary stable strategy, and using Matlab to simulate the situation under different circumstances. The influences of four parameters, including the degree of reputation improvement brought by government regulation, on the evolution path, are analyzed. The study finds : (1) When the benefits of government regulation are greater than those of non-regulation, and the benefits of platform self-discipline are greater than those of non-self-regulation, ideal result (regulation, self-regulation) will appear; (2) When the cost of platform self-regulation is low and the cost of government regulation is high, even if the government does not supervise, the platform will also be self-disciplined; (3) The increased public affirmation of government regulation and the reduction of government regulation costs leads the government to transform from non-regulation to regulation; (4) When the platform obtains positive benefits from society and reduces the cost of self-discipline, its strategy changes from non-self-discipline to self-discipline. These conclusions have guiding significance for the government to optimize the design of regulatory policy.

The traditional frost detection method based on image processing technology is difficult to flexibly and accurately detect the cold chain refrigeration unit in complex production environment. It is also easy to misjudge due to the influence of environmental factors such as illumination and fog. Therefore, a cold chain refrigerating machine frosting detection method based on spiking neural network is designed. Taking the refrigerator image as the input, the dynamic change of the frosting area of the refrigerator evaporator is automatically detected, the abnormalities caused by interference factors such as illumination and fog are corrected, and the double threshold divided by the cumulative value of pulse emission rate is used as the judgment basis of frosting degree. Experiments are carried out on several cold chain refrigerators put into the production environment. The results show that the designed spiking neural network can adaptively detect the dynamic region of the frosting area of the refrigerator evaporator under the actual production environment and the double threshold which is divided accurately can judge the frosting degree of the evaporator. The detection effect is good and the stability is strong, which can provide a reliable basis for the defrosting time of the defrosting strategy of the cold chain refrigeration unit.

Although the Chief Information Officers (CIOs) have the important responsibility for digital innovation in their organizations, it is difficult for them to meet the challenges of the organizational changes involved in relying on traditional structural power and expert roles. Drawing on political skill and issue selling theory, a study is conducted to explore how CIOs can use political skills to drive digital innovation. A theoretical model is presented and empirically tested using survey data collected from a sample of 218 matched pairs of CIOs and top business executives through multilevel regression analysis. The results suggest that: CIO political skill positively affects CIO issue selling effectiveness, which in turn positively affects enterprises digital innovation. CIO issue selling effectiveness fully mediates the positive effect of CIO political skills on enterprises digital innovation. Socio-behavioral integration between business executives positively moderates the relationship between CIO issue selling effectiveness and digital innovation. Accordingly, it is concluded that CIO political skills and issue selling effectiveness are enablers for digital innovation, and that CIO issue selling effectiveness and business executive socio-behavioral integration play a mediating and moderating effect, respectively. The research model proposed and empirically tested in this study bridges the research gap in CIOs effectively driving enterprises' digital innovation and can then provide some practical guidance.

In term of the event-trigger impulsive mechanism, this paper designs a new event triggering function based on the Lyapunov function, and the leader-following consensus of multi-agent systems with randomly occurring uncertainties and randomly occurring nonlinearities is studied. Different from the control method of artificially setting the impulse time sequence, the generation of the impulsive moment depends on the designed triggering function, and when the trigger condition is met, the impulsive control is activated to reduce unnecessary control times and resource consumption. Based on impulsive differential equation theory, algebraic graph theory, and Lyapunov stability theory, the sufficiency conditions that controlled multi-agent systems can achieve the leader-following consensus are given. Meanwhile, the Zeno behavior can be excluded. Finally, the feasibility of the obtained results is verified by a numerical example.

Visual odometry is the cornerstone in the field of SLAM. Monocular visual odometry occupies an important position because of its low cost and less camera calibration, but it has some shortcomings such as scale uncertainty, scale drift, poor robustness, and so on. To solve these problems, based on ORB-SLAM3, we process a monocular visual-inertial navigation odometer with depth features, referred to as DF-VIO (Visual Inertial Odometry Based on Deep Features) , which uses depth features extracted by deep learning network to replace traditional artificial point features, and fuses artificial line features to enhance the robustness of the system in real complex scenes. Besides, the system provides a variety of pose tracking methods. In addition to the method based on the constant speed model and tracking reference keyframe, a pose tracking method based on the predicted repeatability map is also provided, which further improves the accuracy of system pose tracking. Comparative experiments are carried out on the open data set EuRoC, and the average trajectory error is reduced by 25.9% in pure visual mode and 8.6% in visual-inertial mode, which proves that the system proposed in this paper can be more robust in complex scenes.

Low-dose single-photon emission computed tomography (SPECT) imaging can reduce the radiation damage to human bodies caused by radioactive tracers, and hence it is becoming more and more important in clinical practice. In a SPECT system, low-dose imaging can be achieved by acquiring projection data of sparse-view. The sparse-view projection data, if directly reconstructed by conventional iterative reconstruction methods, will inevitably lead to severe ray artifacts in the image domain. Existing clinical reconstruction methods usually introduce specific regularization to the optimization model to suppress ray artifacts. However, this type of methods may not adapt to projection data with various dosage, and the form of regularization heavily depends on prior knowledge. A novel neural network architecture is proposed to learn the mapping from the sparse-view projection data to the full-view projection data. The projection data of missing view angle is synthesized by the proposed neural network to improve the quality of reconstructed images. Numerical experiments show that, compared with the traditional iterative reconstruction method, the SSIM of the reconstructed image is increased by 59%, the NMSE is reduced by 67%, and the PSNR is increased by 2.48 dB. Therefore, the proposed method can better improve the image quality of sparse-view projection data.

Brain tumors are a major problem affecting the health status of the nation. To determine the extent of brain tumor in order to determine the next step in treatment, a puncture biopsy procedure of brain tumor tissue is often required. Magnetic resonance imaging (MRI) is commonly used to detect brain tumors due to its better soft-tissue resolution. Therefore, research related to MRI-compatible robots is necessary. Based on an MRI-compatible hydraulically driven puncture surgery robot, the kinematic model of the robot is derived based on the principle of hydraulic linker, and the dynamic model of the robot is obtained based on the relevant theory of fluid dynamics. In order to realize the accurate control of the designed robot system, the state feedback H_∞ control rate of this hydraulically driven system is designed according to the H_∞ control theory, which enables the robot to track the target signal quickly and stably. Finally, the average positioning accuracy of the designed robot system in x-axis, y-axis, z-axis, pitch-axis and roll-axis are obtained through experimental studies, and they are, respectively, 0.41 mm, 0.6 mm, 0.67 mm, 0.886° and 1.17°. Experimental results verify the performance of robot-assisted positioning puncture needles, and the dynamic model and control method of the robot have been given, which provide certain reference value for the research of the control algorithm of puncture robots.

In order to improve the interaction between go playing and teaching, a drop guidance device for go robot is designed, including visual processing and motion control modules. For the visual processing module, a standard chessboard image extraction method based on multi-scale detection is proposed to improve the stability of chessboard image extraction. A separate chessboard detector is used to detect the pieces in the reflective area of the chessboard, and the detection effect of chess pieces in uneven illumination area is improved. For the motion control module, the high-precision digital actuators and laser guider are used to construct the motion actuator. The kinematics modeling is completed by using the telescopic joint to simulate the laser light path. An error compensation method based on perspective transformation is proposed to realize the mapping of joint variables. The motion end position compensation is implemented through simulation calculation. Finally, the accuracy of vision module and the effectiveness of error compensation method of motion control module are verified experimentally.

Geological disasters caused by surface deformation are common, especially in urban areas, which seriously hinders the sustainable development of urbanization. In order to evaluate the risk of geological disasters caused by urban surface deformation, monitoring and analysis with high temporal and spatial resolution becomes particularly important. Based on Sentinel-1A satellite images from January 2019 to January 2021, SBAS InSAR and PS InSAR technology is used to obtain the time series of surface deformation in Nanchang, and wavelet periodic analysis and grey correlation analysis combined to evaluate the correlation between deformation area and climate environment. The research shows that the central urban area of Nanchang is affected by urban construction and clay layers, showing a large-scale subsidence signal; the agro-ecological areas in the southeast show uplift signals due to groundwater replenishment. Periodic analysis further shows that the surface deformation of Nanchang is affected by the change of precipitation. Combined with external data, this study examines the possibility of subsidence disasters of subway lines in areas with large deformation from multiple perspectives, thus providing reference for disaster prevention and reduction.

A preventive maintenance strategy for key components of ships is advanced based on the delay time theory. This strategy aims at the problem of high maintenance cost and low utilization rate caused by unreasonable maintenance of key parts. It bases on periodic inspection of each component, and divides the maintenance of each component into inspection maintenance and delay maintenance according to the different characteristics of defect and failure stage. And the idea of opportunity maintenance is introduced in the delay phase, to establish an opportunistic maintenance model for the ship’s key components. According to this strategy, the opportunistic maintenance strategy of the key parts under long-term use can be obtained in four steps: Base on component failure data and inspection maintenance records. Estimate model parameters by maximum likelihood estimation. Take detection cycle and delay time as optimization variables. Take the minimum cost ratio as the optimization objective. The result of this research indicates that this strategy can more effectively reduce the cost of ship maintenance, reduce downtime for maintenance, improve the efficiency of the use of the ship, and provide a decision basis and reference for ship maintenance, than the simple test maintenance strategy and traditional preventive maintenance method.

An asymmetric topology suitable for three-phase active power filter is proposed. The topology combines a passive filter with a low-power active filter to decrease the capacity and system losses of the voltage inverter (VSI) . It consists of a three-phase bridge, a two-phase dual-resonant LC passive filter and a phase coupled inductor. By reducing the number of devices, the topology design scheme explores the impact of the changes of APF device parameters, which will affect the compensation effect of APF. Finally, a control strategy based on virtual capacitor voltage compensation is proposed to control the system effectively. Simulation and experiments show that the topology and control strategy proposed are feasible.The scheme of this topology and its control strategy can reduce the number of devices and be used in combination with traditional symmetric control strategies, which is easy to be implemented in engineering.

The carbon emission calculating method of light vehicle is studied, and the relationship between vehicle carbon emission and operating conditions analyzed. Based on the on-board test data of real drive emission (RDE) of vehicles, the carbon emission is represented by CO₂ equivalent CO_2e. It is indicated that the carbon emission rate increases with the increase of vehicle speed and specific power. The Back Propagation (BP) neural network algorithm is used to establish the nonlinear relationship between vehicle carbon emission and speed, acceleration and vehicle specific power (VSP), and calculate the carbon emission factors under three bench test cycle conditions of World Light Vehicle Test Cycle (WLTC), New European Driving Cycle (NEDC) and China Light-duty Vehicle Test Cycle-commercial Car (CLTC-C). It is found that the carbon emission factors under the three bench test cycle conditions are higher than the actual road driving carbon emission factors, among which the carbon emission factor under WLTC is the highest, followed by NEDC and CLTC-C. The reason is that the test conditions with higher acceleration and speed lead to increased carbon emissions.

In order to study the changes of combustion characteristics of domestic gas stoves after mixing natural gas with hydrogen, numerical simulation and thermal experiment research are carried out on the application of natural gas hydrogen-mixing combustion technology in household swirl gas stoves. First, the detailed mechanism GRI-Mech 3.0 is simplified. Through comparison and verification, the error between the simplified mechanism and the original mechanism is less than 1%. Secondly, using simplified mechanism for numerical simulation of pure methane gas stove, the numerical simulation and thermal experimental calculation results are compared and verified. It is found that the maximum error of the average temperature of the combustion zone does not exceed 12%, which confirms the feasibility of the simulation. Finally, a numerical simulation study of methane hydrogen mixing conditions is carried out. The calculation results show that the primary air coefficient of domestic gas stoves gradually increases with the increase of the hydrogen mixing ratio, which has a great influence on the combustion characteristics. After adding 15% of the volume of hydrogen, CO emissions are reduced by about 9%; and the temperature remains basically stable, which confirms the effectiveness of methane hydrogenation technology. The research results have reference value for the popularization and application of natural gas mixed with hydrogen in the combustion of gas stoves.

With excellent mechanical properties and heat resistance, polyimide has been one of the important dielectric materials in microelectronics industry. In recent years, with the rapid development of microelectronics industry and the rise of 5G communication technology, higher requirements are put forward for the dielectric constant and dielectric loss reduction of polyimide in terms of energy consumption requirements and signal reception. How to reduce as much as possible the dielectric constant and dielectric loss and retain the excellent properties of polyimide at the same time is an urgent problem to solve at present. The research and application progress of porous polyimides, polymers filler and inorganic nano filler composite modified polyimides and other composite low dielectric polyimides emerging in recent years are reviewed. How to reduce the dielectric constant of polyimides while maintaining other properties is discussed, and its development is prospected. It will provide a new idea for the design and preparation of new composite low dielectric polyimide materials.

The protective effect of Schisandra chinensis oil on H₂O₂-induced oxidative damage of HaCaT cells is evaluated by 3D skin model. Schisandra chinensis oil was extracted by steam distillation, and its chemical constituents analyzed by GC-MS. And the antioxidant activity of schisandra chinensis oil was evaluated by DPPH and ABTS free radical scavenging experiments, combined with 2D and 3D skin models induced by H₂O₂. The results showed that Schisandra chinensis oil mainly contained 37 compounds including ylenene, β-cimarene and L-α-pinene. Schisandra chinensis oil had good scavenging effect on DPPH and ABTS free radicals, with IC₅₀ of 5.6 mg/mL and 9.4 mg/mL, respectively. What’s more, Schisandra chinensis oil can improve the survival rate of HaCaT cells after H₂O₂ treatment, increasing the epidermal thickness of 3D skin model, decreasing the level of Interleukin-6(IL-6), and up-regulating the activities of SOD and CAT enzymes. The results confirmed that Schisandra chinensis oil had protective effect on H₂O₂-induced oxidative stress injury model of HaCaT cells, which could lay a foundation for the development of anti-aging products of schisandra chinensis oil.