Search

Aiming at the problem of due-date forecasting for make-to-order production environment, this paper proposed a forecasting method namely BDDF, which integrated multi-level workloads such as order accept, order release and workshop scheduling. This method first estimated the workload of "drum-chain" and "non-drum chain" orders at different levels according to whether the order passed through the bottleneck, and then estimated the throughput time of each level by using Little's law to sum and predict the due-date of the order. This paper used the FlexSim modeling and simulation to compare the system performance of BDDF method and two classical due-date prediction methods under different sequencing rules in the pre-shop, dispatching rules in the workshop, and protective capacity for the bottleneck. The results show that the BDDF method is superior to the classical methods in terms of mean tardy rate, mean tardy and standard deviation of lateness. At the same time, pre-shop sequencing rules, scheduling rules and protective capacity all affect the performance of BDDF method. It is necessary to select appropriate value according to the actual situation to better control the order due-date and avoid order delay. [ABSTRACT FROM AUTHOR]

The existing network traffic prediction models have problems such as weak generalization ability and low prediction accuracy. To solve this problem, This paper point out a prediction model combining dynamic diffusion convolution module and convolution interaction module. The dynamic diffusion convolution module can extract the complex spatial and dynamic characteristics of network traffic, while the convolution interaction module can capture the temporal characteristics of the traffic. The organic combination of the two can effectively predict the traffic in the network. This paper verifies the effectiveness of the proposed dynamic diffusion convolutional interaction graph neural network model (DDCIGNN) by comparative experiments with other network traffic prediction models on the flow data of the US Energy Science Network (ESnet). Experimental results show that the root mean square error (RMSE) of the proposed model is optimized by about 13.0% in the best case, which indicates that the model can perform better in network traffic prediction. [ABSTRACT FROM AUTHOR]

Aiming at the shortcoming of salp swarm optimization algorithm such as low accuracy, slow convergence speed and easy to fall into local optimum, this paper proposed an adaptive t-distribution and dynamic weight salp swarm optimization algorithm. Firstly, the leader position update introduced the global search stage formula of butterfly optimization algorithm to enhance the global exploration ability. Secondly, the follower location update introduced an adaptive dynamic weighting factor to strengthen the guiding role of elite individuals, so as to enhance the local development ability. Finally, the adaptive t-distribution mutation strategy mutated the optimal individual in order to avoid the algorithm falling into local optimum. By solving 12 benchmark test functions, and according to comparison results of the mean value, standard deviation, solving success rate, Wilcoxon test and convergence curve, the proposed algorithm was superior to standard salp swarm algorithm, the compared other improved salp swarm algorithm and the compared other swarm intelligence algorithms. The results also show that it has a significant improvement in the optimization accuracy and convergence speed, and has the ability to jump out of local optimum. The experimental results verify the effectiveness of the proposed algorithm by applying it to find the lowest point of denitrification inlet concentration. [ABSTRACT FROM AUTHOR]

In order to accurately predict the short-term passenger flow of urban rail transit for multiple stations, this paper proposed a deep learning model, SAE-ConvLSTM, combining convolutional long short-term memory ( ConvLSTM) and stack autoencoder (SAE) . This paper considered thirteen external factors related to passenger flow, whose features would be extracted by SAE with successive layers and thus obtain more representative features. It proposed ConvLSTM to extract spatiotemporal features of passenger flow, which was combined with the resulting external factors to predict short-term passenger flow of multiple stations simultaneously. And it developed latent action Monte Carlo tree search ( LA-MCTS) to optimize the parameters of SAE. Compared with genetic algorithm( GA), particle swarm optimization ( PSO), simulated annealing algorithm (SA) and tabu search (TS), LA-MCTS performed best in terms of effect and efficiency. This paper conducted extensive experiments. The results show that SAE-ConvLSTM works better than shallow machine learning model-back propagation neural network ( BPNN), support vector regression mode(SVR),autoregressive integrated moving average model ( ARIMA ),and deep learning modellong and short time memory network ( LSTM), convolutional neural network (CNN) and Conv LSTM without external features, ConvLSTM external features without SAE, CNN + LSTM and CNN + LSTM with external features, in terms of root mean square errors ( RMSE), mean absolute errors ( MAE) and mean absolute percentage errors ( MAPE), and the goodness of fit ( R² ) . [ABSTRACT FROM AUTHOR]

Aiming at the shortcomings of the sparrow search algorithm such as slow convergence speed, insufficient optimization accuracy and easy to fall into the local optimum, this paper proposed an improved sparrow search algorithm that combined the sine search strategy and the diversity mutation processing strategy. Through the introduction of a sine search strategy, adaptive adjustment of individual weights improved the convergence speed of the algorithm. Aiming at the problem of excessive individual aggregation, this paper adopted diversity mutation processing, introduced the concept of population aggregation degree in biology and Cauchy mutation to disturb the optimal solution, and improved the possibility of the algorithm escaping from the local optimal. By testing 9 benchmark functions with different characteristics, the test results show that the improved algorithm can converge to the optimal value faster, with better average and standard deviation, indicating that it has better convergence speed, convergence stability, and the ability to escape local optimal values. By applying the improved optimization algorithm to the parameter tuning of the fractional PID controller, the experimental results further verify the effectiveness and feasibility of the improved strategy. [ABSTRACT FROM AUTHOR]

In order to solve the problems of BBO such as insufficient search scope in the early stage and easy to fall into local optimization in the later stage, this paper proposed an improved El-BBO which based on El. Firstly, the algorithm searched for new habitat around the original habitat, it enhanced the diversity of the initialization population. Secondly, the algorithm made improved habitat expansion, it improved the convergence efficiency of the algorithm. Finally, the algorithm used gradient de scent to make quadratic convergence in the field of optimal solution, which improved the convergence accuracy of the algorithm. This paper carried out 50 Monte Carlo experiments on 12 optimized test functions commonly used in CEC2014, the experimental results show that the overall performance of El-BBO is better than the other three intelligent optimization algorithms in terms of optimal fitness value, average fitness value and standard deviation. It shows that El-BBO can improve the ability to find the optimal solution and enhance the search stability. [ABSTRACT FROM AUTHOR]

This paper proposed a two-stage emergency logistics transportation and redistribution strategy for the problems con- sidering the transportation timeliness and the supplies distribution fairness in emergency logistics. Firstly, This paper proposed the capacitated vehicle routing problem with maximum time satisfaction( CVRP-MTS) for emergency vehicle transport schedu- ling and the emergency supplies distribution problem with minimum weighted standard deviation of satisfactions ( EMDP-MWS- DS) based on the time satisfaction, the supplies satisfaction and the comprehensive satisfaction defined in this paper. Then this paper designed a hybrid grey wolf optimization algorithm( HGWOA) to solve the CVRP-MTS in the first stage. The HGWOA adopted the optimal split procedure decoding strategy, and adopted the chaotic search strategy, the 2-0pt and 1-1 local search strategy to improve the search ability of the algorithm. And this paper implemented the chaotic grey wolf optimization algorithm ( CGWOA) to solve the EMDP-MWSDS in the second stage. The experimental results show that in the first stage, overall time satisfaction of the CVRP-MTS is 39. 31 % higher than that of the CVRP model. In the second stage, when 80% of the actual material quantity and 0. 75 for the weight coefficient, the standard deviation of the comprehensive satisfaction degree is 0. 061 9, and the standard deviation of the supplies satisfaction degree is 0. 097 4. The distribution results effectively balance the victims' psychological demands and supplies utilization efficiency. The two-stage emergency logistics transportation and supplies redistribution distribution strategy can obtain high time satisfaction and good supplies distribution results, which prove the effectiveness of the proposed models and algorithms. [ABSTRACT FROM AUTHOR]

In view of the data dependence that is the major drawback of the supervised deep neural network ( DNN )-based denoising models, this paper proposed an unsupervised deep image generation ( UDIG) denoising model. Firstly, it utilized a noise level estimation ( NLE) algorithm to measure the noise level of a given noisy image, the estimated noise level and the noisy image fed into the state-of-the-art denoising model ( i. e., fast and flexible denoising convolutional neural network, FFDNet) to obtain a preliminary denoised image as the input to the UDIG model. Secondly, it chose the encoder-decoder architecture as backbone network, while using the sum of the mean square error among the output image of UDIG model ( i. e . generated image), and the preliminary denoised image, and the given noisy image to define mixed loss function. Then, the regular stochastic gradient descent ( SGD) algorithm optimized the loss function, adjusted the hyper-parameters of the UDIG model to generating a series of generated images. Finally, when the standard deviation of the residual image between the noisy image and generated image approximates the noise level measured by the NLE algorithm, the network iterative process was terminated adaptively, ensuring the image quality of the gene-rated image ( treated as the denoised image) . Extensive experiments show that, the proposed UDIG denoising model has a better performance than other state-of-the-art counterparts with regard to denoising effect. [ABSTRACT FROM AUTHOR]

To address the problem of mismatches produced by traditional stereo matching algorithms when matching low-texture face images,this paper proposed a face stereo matching algorithm based on region growing.This algorithm used the facial feature points extracted by the ensemble of regression trees to divide the face into different regions and limited the disparity search range of different regions,thereby avoided finding matching points on the global scale.Meanwhile,it used the local shape characteristics of the face and the local surface fitting to filter out mismatched seed points and generate a large number of reliable seed points for region growing.Finally,this paper performed qualitative and quantitative experiments on face images acquired in laboratory environment and the FRGC v2.0 face database respectively.Experimental results show that compared with the traditional algorithm,the improved algorithm can reconstruct a more accurate 3 D face model,the root mean square error is within 2 mm after the point cloud registration.In addition,the reconstructions of face images under different lighting,poses and expressions show that the improved stereo matching algorithm has better robustness. [ABSTRACT FROM AUTHOR]

Aiming at the periodic load imbalance caused by user equipment mobility and cell coverage,this paper introduced a cellular network to support high data rate services and intensive deployment,and proposed an adaptive load balancing algorithm for cellular networks considering load estimation.Firstly,the algorithm switched parameters according to the adjustment of overload units and adjacent units.The utilization of resources depended on the signal quality and the traffic demand of the connected user equipment in the long-term evolution(LTE).This paper defined the utilization rate of resource blocks as a measure of unit load,and determined the overload list by using adaptive threshold according to the network load situation.In order to avoid performance oscillation,considering the impact of mobile load on the network,it designed the load balancing process.Finally,it evaluated the performance of the algorithm in different environments by system-level simulation.The simulation results show that the algorithm provides a more balanced load between networks(so that the standard deviation between cells is smaller),and has higher network throughput than previous algorithms. [ABSTRACT FROM AUTHOR]

The traditional matrix decomposition model cannot fully explored the intrinsic relationship between the user and the object in the mean, bias and characteristics. This paper proposed a fitting matrix model to improve the prediction performance by constructing the user and the item matrix to represent the characteristics of the user and the item respectively. The matrix decomposition model had the advantage of accuracy in the field of recommender system, but the gradient descent method, which was the most popular method to train parameters of model, had a slow convergence speed. To resolve the above defects, this paper considered to accelerate the convergence speed using the convergence of quasi Newton method, and named the proposed algorithm as fitting matrix and two orders fusion iterative (FAST) algorithm. The experimental results show that the FAST algorithm is better than the traditional non negative matrix decomposition (NMF), singular value matrix decomposition (SVD), and the regularized singular value matrix decomposition (RSVD). FAST algorithm has a decrease with regard to the mean absolute error (MAE) and the root mean square error (RMSE), and has a significant improvement in the iterative efficiency, which alleviates the problem that the accuracy is difficult to balance with the efficiency of the iteration. [ABSTRACT FROM AUTHOR]

In order to ensure the normal operation of the equipment and to predict the remaining life of the bearing, this paper proposed a life prediction model based on the combination of two-dimensional convolutional neural network and an improved WaveNet. To overcome the gradient vanishing problem of the unoptimized recurrent networks in the process of prediction training, the WaveNet time-series networks structure was introduced into the model. Aiming at the situation that the original WaveNet structure was not suitable for rolling bearing vibration data, the improved WaveNet structure combined with two-dimensional convolutional neural networks was applied to the life prediction of rolling bearing. The model extracted the features of one-dimensional vibration sequence using two-dimensional convolutional networks, and then input the features to the WaveNet to predict the remaining life of the rolling bearing. Compared with the deep recurrent networks, the combined model has higher computational efficiency and more accurate results. Compared with the CNN-WaveNet-0 model, the improved model has more accurate prediction results. Compared with the deep long short-term memory networks model, the root mean square error of the prediction results of this model is reduced by 11.04%, and the scoring function of the prediction results is reduced by 11.34% . [ABSTRACT FROM AUTHOR]

This paper proposed an improved salp swarm algorithm that aimed to solve the problem of low precision and slow convergence speed in solving function optimization problems. For leaders, this algorithm replaced the optimal individual position with weighted centroid to prevent premature aggregation near the optimal individual. It also introduced adaptive inertia weight to the followers to balance the global search and local optimization capabilities of the algorithm. Furthermore, this algorithm contributed to reducing the inter-dimensional interference and improving the diversity of the population by dimensional random difference mutation. The simulation results illustrate that the improved algorithm is better than the standard tympanum group algorithm and other improved algorithms in terms of mean, standard deviation and convergence curve. The results also show that the improved algorithm improves the performance of optimization, and has higher precision as well as faster convergence speed. [ABSTRACT FROM AUTHOR]

Aiming at the problem that the WOA was easy to fall into local optimum and low convergence precision, this paper proposed an improved whale optimization algorithm (IWOA). The algorithm initialized the population by quasi-reverse learning methods and improved the diversity of the population. Then the algorithm modified the linear convergence factor to a nonlinear convergence factor, which was beneficial to balance the global search ability and local development ability. In addition, the algorithm improved the local search ability of the whale optimization algorithm by increasing the adaptive weight and imp roved convergence precision. Finally, the algorithm adjusted the whale optimization algorithm in time by a random differential mutation strategy to avoid falling into the local optimum. It selected nine benchmark functions in the experiment, and iterated all the algorithms 30 times. The improved whale optimization algorithm compared to the original whale optimization algorithm and five improved whale optimization algorithms, the results show that the mean and standard deviation of the algorithm are better than other algorithms, the convergence curve of the algorithm is also superior to most other algorithms. It shows that the imp roved whale optimization algorithm has the best convergence accuracy and algorithm stability, and the convergence speed is significantly faster than most other improved whale optimization algorithms. [ABSTRACT FROM AUTHOR]

For the problem of signal denoising, this paper proposed a time-fractional nonlinear diffusion smoothing method by combining the classical nonlinear diffusion filtering and the fractional derivative. The proposed model used the fractional order time derivative to replace the time derivative in the classical nonlinear diffusion model. Firstly, it developed the numerical algorithm for the time-fractional nonlinear diffusion filtering, and then compared the proposed method with the classical smoothing methods ( nonlinear diffusion filtering, regularization method, Savitzky-Golay method and wavelet method) . The result shows that the time-fractional diffusion filtering method has a greater improvement in signal-to-noise ratio and root mean square error. Compared with the classical smoothing method, the proposed model obviously has a more excellent performance. [ABSTRACT FROM AUTHOR]

The conventional preamble-based channel estimation schemes for filter hank multi-carrier with offset quadrature amplitude modulation (FBMC/OQAM) systems have low estimation accuracy and the use of preamble for channel estimation decreases the system’s spectrum efficiency. This paper proposed a subspace blind channel estimation method for FBMC/OQAM systems. The proposed method distinguished itself from previously preamble based methods by utilizing spatial diversity technique to introduce data redundancy for blind channel estimation, thereby leading to high spectral utilization. Simulation results verify the efficacy of the proposed method in FBMC/OQAM systems and show that the proposed method can provide significantly root mean square error (RMSE) performance improvement than conventional preamble based methods at high SNRs. [ABSTRACT FROM AUTHOR]

In order to detect real-time video streaming in the sender or an intermediate mode and monitor the quality of networks,this paper developed a mobile video satisfaction model based on spatial-temporal complexity in compressed-domain.The model extracted percentage of 4 × 4 block and DC coefficient from I-frame to represent the texture information,and extracted the foreground region from P-frame by contrasting the motion vector of current block and the whole frame,used foreground motion to represent the temporal complexity. Then it established the video satisfaction model for coding distortion and transmission distortion respectively by considering the spatial-temporal complexity. Experimental results show that the R-square between the observed value and the estimated value are 0. 912 and 0. 930,and the root mean square error( RMSE) are 0. 189 and 0. 102. The proposed model has advantages of low computational cost and better performance than previous models,it can be used in the video quality assessment system in real-time. [ABSTRACT FROM AUTHOR]

The change of illumination have a serious effect on pedestrian detection algorithms based on image difference, general adaptive background updating methods is difficult to eliminate the effects of illumination change. This paper proposed a new method of illumination changes elimination based on similarity of histogram distribution to solve this problem. It analyzed similarity of histogram distribution on local area between illumination changed image and normal image. It presented standard deviation of image which could use to measure similarity of histogram distribution. Through image difference experiment between two adjacent frames acquired under light frequency change environment, it shows that the method has a outstanding performance in robustness which can segment the moving targets accurately and restrain the changes of illumination significantly. [ABSTRACT FROM AUTHOR]