针对金融事件存在论元专业性强、分散度高等特点，传统事件抽取方法难以实现精准抽取问题，提出了一种融合段落局部和文档全局特征的事件抽取方法。该方法首先对金融公告文档分段后并行地对所有段落利用Fin-BERT（Financial Bidirectional Encoder Representation from Transformers）预训练模型、卷积神经网络和自注意力机制获取文档局部特征；然后利用Bi-LSTM（Bi-directional Long-Short Term Memory）对整篇文档的语义信息进行学习获取文档全局特征；最后将段落局部特征与文档全局特征融合，输出事件论元和事件类型。在金融公开数据集ChFinAnn上进行的实验结果表明，该方法获得了平均80.2%的F 1 值，优于基线模型，证明了该方法的有效性。
A football competition risk intelligent warning method based on an improved Copula model is proposed to address the issues of large errors between warning values and actual values,and multiple false alarms in football matches.Based on the fuzzy comprehensive evaluation matrix,the evaluation system for football competition risk indicators is determined.The indicator level status is classified,the Copula function is selected,and an improved Copula football competition risk intelligent warning method is constructed to accurately judge football competition risks and reduce risk losses.The experimental results show that the interference suppression of this method is high,maintained above 20 dB,and have high anti-interference ability.It can effectively suppress interference.This method also reduces the error between the warning value and the actual value,reduces the number of false alarms in the warning,and verifies the practicality and feasibility of this method.
Big data clustering processing has the problem of poor clustering effect and long clustering time for different types of data.Therefore,a big data clustering processing method based on the improved PSO-Means(Particle Swarm Optimization Means)clustering algorithm is proposed.The particle swarm optimization algorithm is used to determine the flight time and direction of unit particles during a cluster,preset the selection range of the initial cluster center,and appropriately adjust the inertia weight of unit particles.It eliminates the clustering defects caused by particle oscillation and successfully obtains the clustering center based on large-scale data.Combined with the spanning tree algorithm,the PSO algorithm is optimized from two aspects:sample skewness and centroid skewness.The optimized clustering center is then input into the k-means clustering algorithm to realize the clustering processing of big data.The experimental results show that the proposed method can effectively cluster different types of data,and the clustering time is only 0.3 s,which verifies that the method has good clustering performance and clustering efficiency.
In order to maintain consistency of parameters such as current, voltage, frequency and phase during wind power grid connection, and improve the safety and stability of wind power grid connection, a control parameter identification method for wind power grid connected converters based on optimization algorithm is proposed. Establish a control model for wind power grid connected converters, and change the power control command of the PI regulator based on the power voltage support situation. Set the control conditions of the PI regulator using differential function equations. Calculate the functional relationship in the complex frequency domain, and clarify the logical relationship of the adjustment integral coefficient. The identifiability of control parameters is obtained through the control transfer function, the control output value and characteristic data of parameters are analyzed, and the optimal identification results of control parameters are finally completed. The experimental results show that the proposed method can complete the identification of control parameters in various environments, with small identification errors and high accuracy.
The traditional methods of manually identifying electrocardiogram signals have problems such as high workload and recognition errors.The existing electrocardiogram monitoring equipment still faces drawbacks such as limited recognition types of electrocardiogram signals,low diagnostic accuracy,and excessive reliance on network services.In order to improve the performance of electrocardiogram monitoring systems,ECG(Electrocardiogram Signals)analysis and detection system is designed based on deep learning technology.SENet-LSTM(Squeeze-and-Excitation Networks-Long Short Term Memory)network model is built to realize automatic diagnosis of seven categories of ECG signals.The model is deployed on an intelligent hardware platform which uses ADS1292R as the ECG acquisition module,STM32F103 as the data processing module,and Raspberry PI as the central processing module.The system uses the integrated high-performance microcomputer Raspberry PI for calculation and analysis,and provides users with offline AI(Artificial Intelligence)services.The preciseness of the model can reach 98.44％,and the accuracy can reach 90.00％,realizing the real-time monitoring and accurate classification of ECG,and providing accurate disease diagnosis for patients.
An improved artificial potential field method is proposed to solve the problems of local minimum and unable to reach the target in the path planning of mobile robots.Firstly,in order for the robot to reach the target point when there are obstacles near the target point due to the large repulsive force,a safe distance factor is introduced into the potential field,and this parameter is optimized,so that the robot can maintain a proper distance from the obstacles and reach the target point smoothly.Secondly,in order to solve the local minimum problem,the local minimum discriminant condition is introduced,and the local minimum region is circum-navigated when the condition is triggered,so that the robot can reach the target point smoothly.The simulation results show that the improved algorithm has strong robustness when operating in the map environment with different number of obstacles.The proposed algorithm can make the robot bypass the local minimum area in the U-shaped obstacle environment,and successfully solve the local minimum problem in the mobile robot path planning.
To address the issue of the lack of interoperability and data sharing among different information and application systems on campus,we aim to leverage data integration technology to merge diverse educational data sources.We intend to establish a multi-source,heterogeneous education big data mining and application platform.The platform system will utilize the output of artificial intelligence models and the input from a multi-source,heterogeneous education big data mining engine.It will be based on big data mining techniques to analyze and process multiple data sources,including student records,teaching resources,and social behavior information.This will enable functionalities such as educational sign diagnosis,intelligent learning state comparison,analysis of teaching impact factors,identification of potential issues,and prediction of teaching quality trends.Our goal is to scientifically enhance the quality of personalized campus teaching services,objectively assess the teaching proficiency of individuals and teaching teams,assist in analyzing the strengths and weaknesses of teaching individuals and teams,and provide robust support to decision-makers in managing the education system.
For the problem that LLE(Local Linear Embedding)fails to adequately preserve the structure between neighborhoods in high-dimensional data,a new local linear embedding algorithm is proposed for fused neighborhood distribution properties.The algorithm calculates the neighborhood distribution of each sample data,then calculates the respective nearest neighborhood distribution difference of the KL(Kullback-Leibler)divergence measure between the different neighborhood point and its central sample,and finally optimizes the reconstructed weight coefficient to obtain more accurate low-dimensional motor data.The effectiveness of the algorithm is verified by three evaluations of visualization.Fisher measurement and identification accuracy.
In order to reduce the inspection burden of electric power workers and realize intelligent inspection in substation,the algorithm of substation equipment defect detection is studied.Firstly,the data augmentation method is used to expand the initial dataset and various image processing method is used to generate the dataset with complex illumination environment.Then,the adaptive spatial feature fusion method is used to mitigate the inconsistency of different scale features in the feature pyramid,and the loss function of confidence is changed to Focal loss function to mitigate the imbalance between positive and negative samples.Based on the improved YOLOX-s(You Only Look Once X-s)network model,the algorithm of substation defect detection is designed.Finally,the detection effect of the improved YOLOX-s model is compared with that of other deep learning algorithms.Under the designed data set,the experiment shows that the comprehensive detection effect of the improved YOLOX-s network model is good,and the accuracy and real-time performance is satisfied.
Affected by the stability of the hospital network,the data of paperless office network data is vulnerable to attack.Therefore,an adaptive encryption algorithm based on chaotic sequence is proposed.The key is automatically generated by random numbers,and the chaotic sequence is generated by three-dimensional chaotic system to generate the scrambled and grouped key sequence.Based on the service expectation,a node scheduling algorithm is designed to schedule the nodes of hospital paperless office network to ensure that the key encryption can be scheduled to the appropriate network nodes.Through cloud storage and improved knapsack algorithm,the adaptive encryption of hospital paperless office network is realized.In the paperless office network of a hospital,text data,image data and video data are selected for testing.The test result shows that the encrypted ciphertext presents a digital state,which is not easy to attract the attention of attackers.The mean square error between the ciphertext and plaintext of the three kinds of data is large,up to 258.63.The data correlation of the three kinds of data after encryption is greatly weakened,which shows that the design algorithm can destroy the original correlation of the data and has good network data encryption ability.
At present,due to environmental interference in the process of digital image acquisition and transmission,low-pixel images will appear,resulting in poor image reconstruction effect.For this reason,a digital image super-resolution reconstruction algorithm based on multi-scale residual is proposed.Use bilateral filtering algorithm to complete the dehazing processing of digital images.The brightness feature information and color information of digital images are classfied,and the distance threshold denoising method is used to denoise.To set convolution kernels of multiple sizes.In the process of image feature extraction,digital image features are obtained,and back-projection operations are performed on them.Based on the residual learning idea,the features extracted by the up-sampling and down-sampling processes are connected to realize digital image super-resolution reconstruction.The experimental results show that the proposed algorithm has high structural similarity,high PSNR(Peak Signal-to-Noise Ratio)and good reconstruction effect for image reconstruction.
Most of the existing research on degree prediction in colleges and universities focuses on the construction of performance prediction models,ignoring the importance of degree early warning.Therefore,a degree early warning model based on support vector machine is proposed.A large number of experiments are carried out on the real data of 5 majors,including Broadcast and Television Directing Major,Chinese Language and Literature Major,Chemistry Major,Accounting Major and Mathematics and Applied Mathematics Major,in a university of 2018.The experimental results show that the constructed early warning model has good accuracy and practicality,which can become an important part of improving the teaching quality,and provide practical reference support for teachers to improve the teaching plan and for students to change their learning habits.
In order to improve the torque ripple and flux ripple in the model predictive control system of PMSM(Permanent Magnet Synchronous Motor),a control system scheme is designed by learning the basic structure and control methods of PMSM.The scheme adjusts the duty cycle and voltage vector synchronously.The optimal expected voltage vector and action time at a certain sampling time are selected,and the optimal expected voltage vector and action time at the current sampling time are added to adjust the duty cycle coefficient of the sampling time.The feasibility and effectiveness of this method in improving the control performance of PMSM are verified by comparative analysis of the simulation model.
There are many kinds of raw materials required,and the market price of materials is affected by the price of raw materials.Under such conditions,the price of raw materials has changed,but the market price of materials has not changed.In order to improve the effect of prediction,it is necessary to obtain the time delay of price.This study uses binary density clustering method combined with DTW(Dynamic Time Warping)algorithm,to calculate the similarity between the market prices of commodities over different time intervals and the trends in raw material prices.It has been determined that the market prices of commodities lag behind the fluctuations in raw material prices by 11 weeks.As a result,the market price of cable materials can be predicted based on the trends in raw material prices from 11 weeks ago.This information can assist the procurement department of businesses in formulating rational procurement strategies.
Aiming at the problems of low luminous efficiency and large working bias of AlGaN-based deep-UV(Ultraviolet)LEDs(Light Emitting Diodes),a nitrogen-polar AlGaN-based deep-UV LED device structure with n+-GaN/Al0.4Ga0.6N/p+-GaN tunnel junction is designed.The LED structure is consist of an electron supplying layer n-Al0.65 Ga0.35 N,a multiple quantum wells of Al0.65 Ga0.35 N/Al0.5 Ga0.5 N,a compositionally graded p-AlxGa1-xN,and a n+-GaN/Al0.4Ga0.6N/p+-GaN tunnel junction.The simulation results show that the tunnel junction LED has higher internal quantum efficiency and light output power,and it has a lower turn-on voltage than the reference LED without tunnel junction.The improvement of the optoelectronic characteristics of the tunnel junction LED is attributed to the introduction of the tunnel junction improving the hole injection efficiency of the LED,and improving the current spreading capability of the LED device.The results of this work show that the simulation of carrier transport and optoelectronic characteristics of semiconductor devices through simulation software is helpful deepening the understanding of the physical characteristics of semiconductor devices.If the study of semiconductor device simulation software is added to the learning process of"semiconductor device physics",it will help the cultivation of talents in the semiconductor field.
To monitor changes of indoor carbon dioxide concentrations in real time,a carbon dioxide alarm based on a gas sensor with a STC89C52 microcontroller as the core is designed.When the CO2 concentration in the air exceeds the preset value,the sound and light alarm function can be activated,and the indoor CO2 concentration value can be displayed in real-time.The hardware system includes a carbon dioxide sensor,signal conditioning circuit,analog-to-digital conversion circuit,STC89C52 microcontroller,and acousto-optic alarm unit.The software system includes data acquisition,data processing,alarm logic,and other functional units.The alarm can be activated in time when indoor CO2 concentration exceeds 1.5％.
Due to the influence of complex terrain and canopy structure in forest,the accuracy of snow depth retrieval based on passive microwave remote sensing data is generally low.Based on the representative semi-empirical snow depth retrieval algorithm and combined with meteorological observation data,an optimization algorithm of semi-empirical snow depth retrieval in forest area in Northeast China was established in this paper.In this algorithm,the permittivity of vegetation varies with temperature and the accuracy of snow depth retrieval in forest is greatly improved.Compared with other representative semi-empirical algorithms,the RMSE(Root-Mean-Square Error)of the proposed algorithm is reduced by 2.3 cm,Bias by 3.7 cm on average and correlation(R)improved by 0.11 on average.Compared with the commonly used snow depth retrieval algorithm based on machine learning,the RMSE of the proposed algorithm is reduced by 2.17 cm,Bias by 1.67 cm on average and R improved by 0.22 on average.
In order to improve the experimental teaching of relevant courses in automatic control,the design of a four-tank system is presented.The hardware is composed of general industrial components,while a Matlab programming environment is used for the software to directly control the system.A graphical user interface is also adopted to provide an easy way to run the system.The four-tank system model is built,and its parameters are identified based on a step response experiment.Furthermore,a distributed internal mode controller is designed for the liquid tracking control of the four-tank system.The results show that the four-tank system,using the internal mode controller,has little effect on the water level of the other tank when the water level of one tank changes,and the system has a satisfactory control effect.This experiment could play an important role in the teaching of relevant courses in automatic control.
Currently,SnO2(Tin Dioxide)has become the most commonly used material for the electron transport layer in high-performance PSCs(Perovskite Solar Cells).A strategy for optimizing SnO2 using a small-molecule chelator is proposed to address the problem of agglomeration-prone commercial SnO2 aqueous dispersions and the need to enhance the electrical and surface properties of SnO2 films.The PSCs with the device structure of ITO/SnO2+SC/FA1-xMAxPbI3/Spiro-OMeTAD/Au are prepared by doping the SnO2 transport layer with a low-cost chelator,SC(Sodium Citrate).After the introduction of SC with an optimized concentration,the open-circuit voltage and fill factor of PSCs can reach up to 1.135 V and 78.23％,respectively,with a power conversion efficiency of 21.53％.This represents a significant improvement compared to the devices without the introduction of SC.The characterization of the films and devices revealed that the doping of SC could enhance the electrical and surface properties of the SnO2 films,which in turn improves perovskite crystallization.As a result,defects in the device are reduced,recombination loss is lowered,and charge transport is promoted.
Automatic analysis of lumbar disc herniation requires precise segmentation of the multifidus muscle's fatty infiltration site in spinal MRI(Magnetic Resonance Imaging)images.An attention-based approach for segmenting the multifidus muscle in lumbar disc herniation patients is proposed to address issues including ambiguous boundaries between segmentation targets and adjacent components.The network utilizes an encoder-decoder structure,and the addition of an attention mechanism module to increase the network segmentation accuracy.After feature extraction,an atrous spatial pyramid pooling module is added to combine contextual data improving the performance of the network model.In comparison to the traditional U-Net algorithm,the experimental results demonstrate that this model improves the segmentation accuracy of the fatty infiltrated regions of multifidus muscle by improving the Dice coefficient by 7.8％,Jaccard similarity coefficient by 10.1％,and Hausdorff Distance by 69.5％.
