Flow shop scheduling optimization is one important topic of applying artificial intelligence to modern bus manufacture. The scheduling method is essential for the production efficiency and thus the economic profit. In this paper, we investigate the scheduling problems in a flexible flow shop with setup times. Particularly, the practical constraints of the multi-queue limited buffer are considered in the proposed model. To solve the complex optimization problem, we propose an improved compact genetic algorithm (ICGA) with local dispatching rules. The global optimization adopts the ICGA, and the capability of the algorithm evaluation is improved by mapping the probability model of the compact genetic algorithm to a new one through the probability density function of the Gaussian distribution. In addition, multiple heuristic rules are used to guide the assignment process. Specifically, the rules include max queue buffer capacity remaining (MQBCR) and shortest setup time (SST), which can improve the local dispatching process for the multi-queue limited buffer. We evaluate our method through the real data from a bus manufacture production line. The results show that the proposed ICGA with local dispatching rules and is very efficient and outperforms other existing methods.
Accurately predicting the storage area of prefabricated components facilitates transshipment scheduling and prevents the waste of storage space. Due to the influence of numerous factors, precise prediction remains challenging. Currently, limited research has addressed the prediction of storage areas for prefabricated components, and effective solutions are lacking. To address this issue, a GRU model with an attention mechanism based on ensemble learning was proposed. The model employed the Bo-Bi-ATT-GRU approach to address the time series prediction of storage areas. A Bayesian optimization algorithm was utilized to enhance parameter tuning and training efficiency, while an ensemble learning framework improved model stability. In this study, a port container dataset was used for experimentation, with root mean square error (RMSE) and mean absolute percentage error (MAPE) as evaluation metrics. Compared with the GM model, the R2 of the proposed model improved by 3.38%. Experimental results demonstrated that the ensemble learning-based prediction model offered superior performance in forecasting the storage area of prefabricated components.
1 November 2021 in Journal of Physics: Conference Series
Aiming at the problem that the exiting human skeleton-based action recognition methods cannot fully extract the relevant information before and after the action, resulting in low utilization efficiency of skeleton points, we propose a two-layer LSTM (long short term memory) network with attention mechanism. The network has two layers, the first LSTM network is used for skeleton coding and initialization of system storage units and the second LSTM network integrates attention mechanism to further process the data of the first layer network. An algorithm is designed to assign different weights to skeleton points according to the importance of human body, which greatly increases the recognition accuracy. Action classification is accomplished by multiple support vector machines. Through training and testing, the average recognition rate of 98.5% is achieved on KTH dataset. The experimental result shows that the proposed method is effective in human behavior recognition.
The traffic information forecasting method for route guidance in large traffic network, based on artificial neural network is studied and the time-varied road weight matrixes is constructed, in order to solve the problem of low forecasting rate in traditional and static road weight. An evolution algorithm (EA) for optimal route choice is presented. The corresponding genetic operator, mutation operator and the refresh way for population are proposed. A rectangle restricted searching area (RRSA) method which can reduce the searching area of EA is presented. The problem of bad real-time and astringency of EA for optimal route computing in large traffic network is solved using RRSA. Simulation results have shown that good accuracy and real-time characteristics are got for route guidance in large traffic network.
In order to solve Re-entrant Hybrid Flowshop (RHFS) scheduling problems and establish simulations and processing models, this paper uses Wolf Pack Algorithm (WPA) as global optimization. For local assignment, it takes minimum remaining time rule. Scouting behaviors of wolf are changed in former optimization by means of levy flight, extending searching ranges and increasing rapidity of convergence. When it comes to local extremum of WPA, dynamic regenerating individuals with high similarity adds diversity. Hanming distance is used to judge individual similarity for increased quality of individuals, enhanced search performance of the algorithm in solution space and promoted evolutionary vitality.A painting workshop in a bus manufacture enterprise owns typical features of re-entrant hybrid flowshop. Regarding it as the algorithm applied target, this paper focus on resolving this problem with LDWPA (Dynamic wolf pack algorithm based on Levy Flight). Results show that LDWPA can solve re-entrant hybrid flowshop scheduling problems effectively.
1 July 2007 in Journal of Highway and Transportation Research and Development (English Edition)