Optimization of College English Classroom Teaching Efficiency by Deep Learning SDD Algorithm

Computational Intelligence and Neuroscience, Год журнала: 2022, Номер 2022, С. 1 - 10

Опубликована: Янв. 21, 2022

In order to improve the teaching efficiency of English teachers in classroom teaching, target detection algorithm deep learning and monitoring information from are used, Single Shot MultiBox Detector (SSD) is optimized, optimized Mobilenet-Single (Mobilenet-SSD) designed. After analyzing Mobilenet-SSD algorithm, it recognized that has shortcomings large amount basic network parameters poor small detection. The deficiencies following partThrough related experiments student behaviour analysis, average accuracy reached 82.13%, speed 23.5 fps (frames per second). Through experiments, achieved 81.11% detecting students’ writing behaviour. This proves proposed improved recognition without changing operation traditional algorithm. designed more advantages compared with previous algorithms. improves which beneficial provide modern technical support for understand status students strong practical significance improving teaching.

Язык: Английский

---

Fractional-Order PIλDμ Controller Using Adaptive Neural Fuzzy Model for Course Control of Underactuated Ships

Applied Sciences, Год журнала: 2022, Номер 12(11), С. 5604 - 5604

Опубликована: Май 31, 2022

For the uncertainty caused by time-varying modeling parameters with sailing speed in course control of underactuated ships, a novel identification method based on an adaptive neural fuzzy model (ANFM) is proposed to approximate inverse dynamic characteristics ship this paper. This adjusts both its own structure and as it learns, able automatically partition input space, determine number membership functions rules. The trained ANFM used controller, parallel fractional-order PIλDμ controller for ships. Meanwhile, sine wave curve sawtooth are considered learning samples ANFM, respectively, dynamics simulation experiments carried out. Two different structures obtained, which connected respectively ship. results show that can effectively overcome influence parameters, track desired quickly effectively, has good effect. Finally, comparative four controllers out, FO using advantages small overshoot, short adjustment time, precise control.

Язык: Английский

---

A Study of the Spatial Effects of Financial Support and Economic Growth under Optimal Control of Nonlinear Generalized Complex Systems

Discrete Dynamics in Nature and Society, Год журнала: 2022, Номер 2022(1)

Опубликована: Янв. 1, 2022

This paper provides an in‐depth analysis and study of the spatial effects financial support economic growth with help nonlinear generalized complex systems. Taking industrial sector as research object combining relevant contents neoclassical investment theory, information economics, institutional this clearly defines argues that main feature current policy is constraint rather than inhibition based on understanding theoretical connotation rationality and, a premise, further analyzes causing excessive capital mismatch in corporate sector. It mechanism role policies overinvestment conducts empirical tests from three perspectives measuring efficiency, output efficiency investment, allocation industry capital, behavior microenterprises, finally puts forward recommendations conjunction evaluation policies. selects dimensions system, namely, structure, scale, studies adaptability between these development real economy, respectively, then uses different methods to analyze dynamic economy system explores way economy. medium micro basis new evidence for importance reform also opens up space exploring exit path constraints using interest rate marketization general grip reasonably guide resources achieve transformation upgrading sustainable healthy through supporting high‐quality more interprovincial level data analysis, so it comprehensive detailed previous scholars’ studies. The examination

Язык: Английский

---

[Retracted] Construction of Financial Management Early Warning Model Based on Improved Ant Colony Neural Network

Computational Intelligence and Neuroscience, Год журнала: 2021, Номер 2021(1)

Опубликована: Янв. 1, 2021

With the advent of era economic globalization, world capital market is also facing financial risks. It necessary to have a corresponding management early warning model reduce losses. This paper uses combination ant colony algorithm and neural network build improved by model. By setting relevant assumptions, statements annual report texts are predicted analyzed compared with original static data forecasting Compared traditional methods, time series sequencing analysis used in this makes result prediction more accurate. allows one year’s be predict for next two years. research can provide reference optimization system.

Язык: Английский

---

A Novel Fault Feature Extraction Method for Bearing Rolling Elements Using Optimized Signal Processing Method

Applied Sciences, Год журнала: 2021, Номер 11(19), С. 9095 - 9095

Опубликована: Сен. 29, 2021

A rolling element signal has a long transmission path in the acquisition process. The fault feature of is more difficult to be extracted. Therefore, novel weak extraction method using optimized variational mode decomposition with kurtosis mean (KMVMD) and maximum correlated deconvolution based on power spectrum entropy grid search (PGMCKD), namely KMVMD-PGMCKD, proposed. In proposed KMVMD-PGMCKD method, VMD Then an adaptive parameter selection for MCKD, PGMCKD, determine period T filter order L. complementary advantages KMVMD PGMCKD are integrated construct model (KMVMD-PGMCKD). Finally, employed deal obtained by effectively implement extraction. Bearing signals Case Western Reserve University actual data selected prove validity KMVMD-PGMCKD. experiment results show that can extract features bearing elements accurately diagnose faults under variable working conditions.

Язык: Английский

---