Abstract In practical engineering, the working conditions of gearbox are complex and variable. varying conditions, performance intelligent fault diagnosis model is degraded because limited valid samples large data distribution differences signals. Based on these issues, this research proposes a method integrated with lightweight channel attention mechanism, further realizes cross-component transfer learning. First, time–frequency original signals obtained by wavelet transform. It could...

According to the perishable characteristics of refrigerated food and objective minimizing total cost, mathematical optimization model cold chain logistics distribution center location problem is established by introducing such constraints as freshness time window. In order solve problems slow convergence easy fall into local optimal solution in process traditional wolf colony optimization, an immune hybrid algorithm proposed center. this algorithm, idea vaccination introduced algorithm. By...

For most rotating mechanical transmission systems, condition monitoring and fault diagnosis of the gearbox are great significance to avoid accidents maintain stability in operation. To strengthen comprehensiveness feature extraction improve utilization rate signals accurately identify different operating states a gearbox, model combining Gramian angular field (GAF) CSKD-ResNeXt (channel shuffle kernel decomposed ResNeXt) was proposed. The original one-dimensional vibration signal converted...

Under different working conditions of gearbox, the feature extraction fault signals is difficult, and large difference in data distribution affects diagnosis results.Based on problems, research proposes a method based improved MobileNetV3 network transfer learning (TL-Pro-MobilenetV3 network).Three timefrequency analysis methods are used to obtain time-frequency distribution.Among them, short time Fourier transform (STFT) combined with Pro-MobilenetV3 takes shortest has highest...

Gearbox fault diagnosis based on traditional deep learning often needs a large number of samples. However, the gearbox samples are limited in practical engineering, which could lead to poor performance. Based above problems, this paper proposes method Gramian angular field (GAF) and TLCA-MobileNetV3 achieve fast accurate sample recognition under varying working conditions, further cross-component within gearbox. First, 1D signals converted into 2D images through GAF. Second, lightweight...

Abstract Large-scale use of electric vehicles will greatly increase the traffic pressure on urban road network. Therefore, planning charging stations for considering demand and transportation network is particularly important coordinated development intelligent transportation. Under condition bounded rationality, this paper considers such factors as travel utility perception difference between users fuel vehicles, time-varying flow, location service level stations. On basis, combining...

(Purpose/Significance). This paper aims at the problems of inaccurate recommendation effect caused by data sparseness and cold start in traditional collaborative filtering-based book personalized algorithm. So this proposes a filtering algorithm which improves similarity solution method filling missing data. (Method/Process). By considering influence user’s common rating collection on calculation, average value all books is used as threshold, weight introduced into calculation. As for...

Abstract Aiming at the nonlinear and non-stationarity of gearbox fault signals confusion among different categories, a gear diagnosis method combining variational mode decomposition, reconstruction ResNeXt is proposed in this paper. In paper, parameter K VMD determined according to changing trend sample entropy (SE), modal components are obtained after effective extracted reconstructed Pearson autocorrelation coefficient, so as remove redundant information from original signal. Then signal...

This study analyses consumer last-mile delivery service choice behaviour by developing a cross-nested logit model (CNL); we then compare the analytical results with three nested models (NL). The parameters are estimated using data from questionnaire collected consumers residing in Beijing, Shanghai, Tianjin, Guangdong, Zhejiang, Jiangsu, and Shandong. direct elasticities cross-elasticities calculated to assess change probability of each alternative caused utility variables. Parameter...

Because of its long running time, complex working environment, and for other reasons, a gear is prone to failure, early failure difficult detect by direct observation; therefore, fault diagnosis gears very necessary. Neural network algorithms have been widely used realize diagnosis, but the structure neural model complicated, training time not easy converge. To solve above problems combine advantages ResNeXt50 in extraction image features, this paper proposes gearbox detection method that...

To investigate the static and dynamic trends of scientific research efficiency different types universities in China during period 2016 -2020.Methods: Based on data six nationwide from to 2020, which are classified by Compendium Science Technology Information Higher Education Institutions published Department Ministry Education.We selected full-time personnel (person-years) development Internal expenditure for year science technology funds (thousand yuan) as input evaluation indexes, number...

Purpose: This paper is committed to design a logistics industry development policy model based on system dynamic simulate the measures which promote region economic and efficiency. The interaction between logistic economy needs be investigated influence degree of efficiency affected by identified too. Design/methodology/approach: In order achieve objective, it makes analysis from perspective divide into growth subsystem, demand subsystem supply subsystem. Then hypothesis boundaries are...

This paper aims at studying the changes in number of pedestrians who go across road "Chinese ways" to discover evolution laws and behavior influential factors. The model including crossing analysis without supervising are established, profit matrix is acquired by intercepting scene style crossing". influence variety decision-making parameters on discussed. effect revenue loss ordinary pedestrian adventure illegal pedestrian, changing initial condition results shown numerical simulation. It...