• 期刊

Construction Research on Sports Training Simulation System Based on Computer Modeling

During the process of traditional training on athletes' training mode, due to the lack of comparable training data and the calculation model with higher calculation efficiency, the feasibility of training plan cannot be assessed accurately. Therefore, the sports training simulation system is designed in this research. This system consists of the human-computer communication system, object library, model library, knowledge base, intelligent control system and analytical interpretation system. By collecting relevant training plan indicators of athletes, such as the past training performance data of athletes, the training status data of athletes, time data indicating athletes' training indicators and physical index data of athletes etc., the preferred plan for the comprehensive evaluation on training mode can be further improved, and the suggestions about optimizing the training content can be provided through simulation system, thus the data support can be provided to the decision maker of the training plan.

• 期刊

Improved Faster RCNN Approach for Vehicles and Pedestrian Detection

The detection of vehicles and pedestrians is very important in automatic driving and intelligent assi-sted driving. Traditional detection methods require artificial design features for extraction, which has the problem of poor robustness. Faster-RCNN object detection algorithm is applied to vehicle and pedestrian det- ection. The K-Means clustering algorithm is used to cluster the ground truth to determine the length-width ratio of the anchor frame. Design a feature fusion module to improve the algorithm's ability to use shallow features. Compared with the original Faster-RCNN algorithm, MAP improved by 4.78% in comparison test of KITTI data set, and the detection effect of small targets was improved to some extent.

• 期刊

Research on Multimodal Transport Path Optimization with Mixed Time Windows Constraints

In the study of multimodal transport, the mixed time window is one of the important factors in the optimization of multimodal transport routes. In order to optimize the transportation route and select the mode of the transport multimodal transport network, it is more scientific and reasonable. In this paper, the multimodal transport model multi-objective optimal path established with mixed time window constraints and total costs and carbon emissions as the goals is established. To reduce transportation costs, respond to the theme of ＂low-carbon logistics＂, reduce carbon emissions, optimize transportation routes, and provide a basis for decision-makers to choose a reasonable transportation solution.

• 期刊

Short-term Forecast of Tubarao-Ningbo Iron Ore Freight Rate Based on ARMA Model

Impact of iron ore freight rate fluctuation on mining industry, steel enterprises and ship chartering are important. Using ARIMA and taking the iron ore freight rate from port Tubarao to port Ningbo as an example, with the help of Eviews 10, the historical time series data from January 2005 to October 2019 was preprocessed, modeled and tested, then obtain the ARMA (1,1), a freight rate model with good fitting effect. And the iron ore freight rate of the route was predicted in a short term, the results shows that half year rate will show a slow growth trend in the range of $21.29-23.35 per ton.

• 期刊

Evolution of a Preparation Route to Sulfamoylamino Intermediates of Cephalosporin

Development of a convergent preparation route to sulfamoylamino intermediates of cephalosporin is described.The target molecule is assembled through an azabicyclo [2.2.1] heptane intermediate followed by condensation reaction with sulfamoylamino methylpiperidine derivatives. The preparation of azabicyclo [2.2.1] heptane intermediate by a intramolecular condensation reaction of (2S,4R)-4-Hydroxypyrrolidine-2-carboxylic acid. This avoids handling of column chromatography and l improves the overall yield and efficiency of the route.

• 期刊

Fuzzy PID Control of Offshore Platform Multi-point Mooringsystem

Withthreemodelanchors and one real anchor to simulate the positioning equipment of semi-submersible unit, it establishes the dynamic models of anchor and anchor chain as well as the platform system mathematic model. Automatic positioning algorithm is used to control the unit move. The application of convert, PLC and industrial Ethernet technology improves the automatic level, work effect and reliability, which is significant to domestic development of semi-submersible unit.

• 期刊

Research Progress of Soil Microbial Species Diversity in China

Microorganisms are the most widely distributed life forms, which are distributed to almost all habitats on earth. They can use a variety of organic compounds and inorganic salts as energy to live in cold polar regions, hot springs up to 100 ℃ or extreme environments with high salinity and alkalinity under aerobic or anaerobic conditions. Microorganisms are the second largest group of organisms on the earth after insects, and their species resources are extremely rich. The study of microbial diversity is the main component of the whole biodiversity study. Microbial diversity mainly studies the population, abundance, distribution uniformity, structural changes and functional diversity of microbial communities in the environment . Microbial diversity can be divided into genetic diversity, functional diversity, structural diversity and species diversity, among which species diversity is the most basic content in soil microbial diversity research . The study included all the bacteria, archaea and fungi in the soil.

• 期刊

Sessions Prediction for Open Educational Resources Based on Deep Learning

Open Educational Resources (OER) are increasingly welcomed by people as the demand for education. However, how to make full use of them and make learners study efficiently could be some tackle problems to be solved. In this paper, we start data processing from Yale Open Courses and propose an efficient deep learning based model capturing context similarity by using word embeddings and LSTM to predict the most relevant sessions from any text paragraph from courses so that help to recommend courses or generate learning roads for OER users. The results of the experiment indicate that the model shows high performance on OER courses dataset in the experiment and also provides ideas for sessions' prediction and recommendation.

• 期刊

Extension Based on a Limit Problem

This paper presents five solutions to a limit problem and its extensions. By comparing and analyzing the advantages and disadvantages of several solutions, an effective method for this kind of limit calculation is given.

• 期刊

End-to-end Learning Based Behaviror Clone

In recent years, although the topic of unmanned driving continues, it is still a challenge to realize unmanned driving in an unknown environment. In order to realize auto driving in an unknown environment, FCN-BiLSTM framework was proposed and end-to-end training was adopted. Road images collected by the simulator were input into the network, and the data of the steering Angle of the car was output, with the deep learning network in the middle, so as to control the car simulator and realize the smooth driving of the car. BiLSTM network is composed of forward LSTM network and backward LSTM network, which is used to deal with the dependence between before and after images. Experimental results show that in Udacity open source unmanned driving simulator, the combination of convolutional neural network and bidirectional long and short-term memory neural network provides an effective scheme for the control of unmanned driving.