摘要:
This paper presents a two-stage multi-objective evolutionary algorithm based on classified population (TSCEA) to solve vehicle routing problem with time windows (VRPTW). It is a well-known NP-hard discrete optimization problem with three objectives: to minimize the total distance cost, to minimize the number of vehicles, and to optimize the balance of routes within a limited time. For TSCEA, there are two stages: In the first stage, a population is explored using the proposed algorithm and then classified according to the number of vehicles, we call this process population classification; In the second stage, Pareto solution set of tri-objective VRPTW is obtained by optimizing the classified population again. The advantages of classified population structure are that for the first stage, this population that the number of vehicles of each individual is in this range composed of the upper and lower bounds of vehicles can be classified as different small populations with the same number of vehicles. Due to the evolution of small population, Pareto solution set with better extensibility can be searched. For the second one, it can reduce the dimension of tri-objective function, that is, three objective functions can be reduced to two objective functions because one of them has been identified in the first stage. Moreover, to resolve the nonlinear discrete problems, the computational approach of crowding degree is modified. The paper chooses Solomon benchmark instances as testing sets and the simulated results show that TSCEA outperforms the compared algorithms in terms of quality or extension, which verified the feasibility of the algorithm in solving tri-objective VRPTW.
期刊:
JOURNAL OF INTERNET TECHNOLOGY,2021年22(2):239-255 ISSN:1607-9264
作者机构:
[Zhang, Junjie; Zhang, Cong] Wuhan Polytech Univ, Sch Math & Comp Sci, Wuhan, Peoples R China.;[Chien, Wei-Che] Natl Dong Hwa Univ, Dept Comp Sci & Informat Engn, Shoufeng Township, Hualien County, Taiwan.
关键词:
Deep reinforcement learning;Value function;Policy gradient;Sparse reward
摘要:
The deep reinforcement learning value has received a lot of attention from researchers since it was proposed. It combines the data representation capability of deep learning and the self-learning capability of reinforcement learning to give agents the ability to make direct action decisions on raw data. Deep reinforcement learning continuously optimizes the control strategy by using value function approximation and strategy search methods, ultimately resulting in an agent with a higher level of understanding of the target task. This paper provides a systematic description and summary of the corresponding improvements of these two types of classical method machines. First, this paper briefly describes the basic algorithms of classical deep reinforcement learning, including the Monte Carlo algorithm, the Q-Learning algorithm, and the most primitive deep Q network. Then the machine improvement method of deep reinforcement learning method based on value function and strategy gradient is introduced. And then the applications of deep reinforcement learning in robot control, algorithm parameter optimization and other fields are outlined. Finally, the future of deep reinforcement learning is envisioned based on the current limitations of deep reinforcement learning.
期刊:
DISCRETE AND CONTINUOUS DYNAMICAL SYSTEMS,2021年41(4):1971-1999 ISSN:1078-0947
通讯作者:
Ruan, Lizhi
作者机构:
[Fan, Lili] Wuhan Polytech Univ, Sch Math & Comp Sci, Wuhan 430023, Peoples R China.;[Ruan, Lizhi] Cent China Normal Univ, Hubei Key Lab Math Phys, Sch Math & Stat, Wuhan 430079, Peoples R China.;[Xiang, Wei] City Univ Hong Kong, Dept Math, Kowloon, Tat Chee Ave, Hong Kong, Peoples R China.
通讯机构:
[Ruan, Lizhi] C;Cent China Normal Univ, Hubei Key Lab Math Phys, Sch Math & Stat, Wuhan 430079, Peoples R China.
摘要:
This paper is devoted to the study of the inflow problem governed by the radiative Euler equations in the one-dimensional half space. We establish the unique global-in-time existence and the asymptotic stability of the viscous contact discontinuity solution. It is different from the case involved with the rarefaction wave for the inflow problem in our previous work [6], since the rarefaction wave is a nonlinear expansive wave, while the contact discontinuity wave is a linearly degenerate diffusive wave. So we need to take good advantage of properties of the viscous contact discontinuity wave instead. Moreover, series of tricky argument on the boundary is done carefully based on the construction and the properties of the viscous contact discontinuity wave for the radiative Euler equations. Our result shows that radiation contributes to the stabilization effect for the supersonic inflow problem.
期刊:
Optimal Control Applications and Methods,2021年42(6):1762-1774 ISSN:0143-2087
通讯作者:
Jiemei Zhao
作者机构:
[Zhao, Jiemei] Wuhan Polytech Univ, Sch Math & Comp Sci, Wuhan 430023, Hubei, Peoples R China.
通讯机构:
[Jiemei Zhao] S;School of Mathematics and Computer Science, Wuhan Polytechnic University, Wuhan, China
关键词:
hybrid time-delays;linear discrete-time systems;reachable set estimation;Wirtinger-based summation inequality
摘要:
This article investigates the problem of reachable set estimation for linear discrete-time systems (LDSs) with both discrete and distributed delays as well as bounded disturbance inputs. The purpose is to determine a bounded set in which all the LDSs states are in the set. By exploiting Wirtinger-based summation inequality and linear matrix inequalities techniques, a sufficient criterion is established which guarantees the states of LDSs are bounded by the ellipsoid. An improved Lyapunov-Krasovskii (L-K) functional is addressed and the L-K functional matrices are not required all to be positive definite. Finally, the effectiveness and superiority of the proposed methods are substantiated by two numerical examples.
摘要:
Hyperspectral technology is used to obtain spectral and spatial information of samples simultaneously and demonstrates significant potential for use in seed purity identification. However, it has certain limitations, such as high acquisition cost and massive redundant information. This study integrates the advantages of the sparse feature of the least absolute shrinkage and selection operator (LASSO) algorithm and the classification feature of the logistic regression model (LRM). We propose a hyperspectral rice seed purity identification method based on the LASSO logistic regression model (LLRM). The feasibility of using LLRM for the selection of feature wavelength bands and seed purity identification are discussed using four types of rice seeds as research objects. The results of 13 different adulteration cases revealed that the value of the regularisation parameter was different in each case. The recognition accuracy of LLRM and average recognition accuracy were 91.67-100% and 98.47%, respectively. Furthermore, the recognition accuracy of full-band LRM was 71.60-100%. However, the average recognition accuracy was merely 89.63%. These results indicate that LLRM can select the feature wavelength bands stably and improve the recognition accuracy of rice seeds, demonstrating the feasibility of developing a hyperspectral technology with LLRM for seed purity identification.
作者机构:
[He Z.; Hu X.; Zhou K.; Shu H.] College of Mathematics and Computer Science, Wuhan Polytechnic University, Wuhan, 430023, China;[Zhou J.] College of Food Science and Engineering, Wuhan Polytechnic University, Wuhan, 430023, China;[Li G.] Enterprise-School Joint Innovation Center, Qianjiang Jujin Rice Industry Co., Ltd., Hubei, China
会议名称:
15th International Conference on Bio-Inspired Computing: Theories and Applications, BIC-TA 2020
会议时间:
23 October 2020 through 25 October 2020
关键词:
Data analysis;Grain transportation;Harmony search algorithm;Multi-objective optimization;Neighborhood search
作者机构:
[Zhao, Jiemei] Wuhan Polytech Univ, Sch Math & Comp Sci, Wuhan 430023, Peoples R China.
通讯机构:
[Zhao, Jiemei] W;Wuhan Polytech Univ, Sch Math & Comp Sci, Wuhan 430023, Peoples R China.
摘要:
Dear editor,
As a consequence of symmetry arguments,the memristor was predicted by Chua[1].As the fourth basic circuit element,its memory characteristic and nanometer dimen-sions are devoid of resistors,capacitors,and inductors.In the field of the dynamical behavior analysis for memristive neural networks(MNNs),information exchange and signal transmission among different neurons are time-varying ac-tivities and discrete time delays are frequently supposed to be bounded,which implies that the current state of a neuron depend only on a part of its history.Actually,the current behavior of a neuron depends upon all its historical infor-mation.Consequently,discrete time delays in MNNs should be assumed to be time-varying and unbounded,which can exhibit the characteristics of the neurons in human brains.Many outstanding achievements on MNNs have already been investigated,but the discrete time delays of the in-vestigated MNNs were all assumed to be bounded[2-4].
作者机构:
[Liu, Zhenqi] University of Sheffield, Automatic Control and Systems Engineering, South Yorkshire, Sheffield, United Kingdom;[Liu, Changhua; Chen, Jiaxi] School of Mathematics and Computer Science, Wuhan Polytechnic University, Hubei, Wuhan, China
会议名称:
5th International Conference on Computer Science and Application Engineering, CSAE 2021
期刊:
Mathematical Problems in Engineering,2021年2021 ISSN:1024-123X
作者机构:
[Yang, M. J.; Sun, C. Q.; Liu, R. F.] Wuhan Polytech Univ, Sch Math & Comp Sci, Wuhan 430023, Peoples R China.;[Zeng, S.] Wuhan Polytech Univ, Sch Econ & Management, Wuhan 430023, Peoples R China.
摘要:
The research about online monitoring and leakage automatic location of water distribution networks (WDN) has a wide range of applications that include water resource protection, monitoring, and allocation. Variational mode decomposition (VMD) and cross-correlation (CC) based leakage location is a popular and effective method in WDN. However, the value of K intrinsic mode functions (IMFs) based on VMD decomposition needs to be determined artificially, which affects the separation effect of signal frequency band characteristics directly. Hence, this work proposes an adaptive method to determine the parameter K of leakage vibration signal's IMFs, which will be applied to automatic leakage location in WDN. Firstly, the number of saddle points in the frequency domain envelope of the sampled signal in different step sizes is calculated. The parameter K is determined according to the curvature change of the number of saddle points and the sampled signal. Finally, the selective IMFs are reconstituted into a new signal, which can determine a leak position using CC based time-delay estimation (TDE). To verify the effectiveness of the proposed algorithm, the different methods based on EMD and Fast ICA are compared. The experimental results demonstrate that the proposed parameter K value adaptive VMD (KVA-VMD) decomposition method is more suitable for leakage location in WDN.