Multi-task and multi-stage commodity sorting algorithm for distributed e-commerce logistics system DOI Creative Commons
Gao Li,

Heyu Yang,

Kai Gu

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: May 9, 2024

Abstract With the increasing scale of e-commerce and logistics, factors such as types, quantities, volumes goods sorted in logistics have made sorting process more difficult. A single serial workshop method cannot meet growing business needs, with low speed inability to make reasonable use machine resources. This article proposes a distributed system commodity algorithm, which sorts multiple types through machines divides task into stages for processing. We present Distributed Dynamic Programming Memetic System (DDPMS): (1) Parallel architecture realizes multi-stage parallel acceleration, dynamic programming algorithm plans execution order tasks reduce total time. (2) The meme first initializes scheduling scheme an improved dragonfly mutation operator prevents solution from falling local optima. (3) And reinforcement search optimization based on fuzzy Q-Learning was designed, using Softmax Green strategy cope complex resource sizes quantities. Improve utilization, time, ensure load balancing. Through experimental verification, efficiency technology has been by 2 4 times compared traditional algorithms; By comparing Dragonfly different numbers instances, converges optimal baseline algorithm. Fuzzy also advantages terms ability efficiency.

Language: Английский

Reinforcement Learning-Driven Proximal Policy Optimization-Based Voltage Control for PV and WT Integrated Power System DOI
Anis Ur Rehman, Zia Ullah,

Hasan Saeed Qazi

et al.

Renewable Energy, Journal Year: 2024, Volume and Issue: 227, P. 120590 - 120590

Published: May 2, 2024

Language: Английский

Citations

9
Improving operations through a lean AI paradigm: a view to an AI-aided lean manufacturing via versatile convolutional neural network DOI
Mohammad Shahin,

Mazdak Maghanaki,

Ali Hosseinzadeh

et al.

The International Journal of Advanced Manufacturing Technology, Journal Year: 2024, Volume and Issue: 133(11-12), P. 5343 - 5419

Published: July 2, 2024

Language: Английский

Citations

9
Multi-objective particle swarm optimization for optimal scheduling of household microgrids DOI Creative Commons
Yu Huang, Gengsheng He,

Zengxin Pu

et al.

Frontiers in Energy Research, Journal Year: 2024, Volume and Issue: 11

Published: Jan. 16, 2024

Addressing the challenge of household loads and concentrated power consumption electric vehicles during periods low electricity prices is critical to mitigate impacts on utility grid. In this study, we propose a multi-objective particle swarm algorithm-based optimal scheduling method for microgrids. A microgrid optimization model formulated, taking into account time-sharing tariffs users’ travel patterns with vehicles. The focuses optimizing daily costs minimizing grid-side energy supply variances. Specifically, mathematical incorporates actual input output each distributed source within as variables. Furthermore, it integrates an analysis capacity variations storage batteries vehicle batteries. Through arithmetic simulation Pareto solution set, identifies that effectively mitigates fluctuations in side. Simulation results confirm effectiveness strategy reducing costs. proposed approach not only improves overall quality but also demonstrates its economic practical feasibility, highlighting potential broader application impact.

Language: Английский

Citations

5
Electricity–gas multi-agent planning method considering users’ comprehensive energy consumption behavior DOI Creative Commons
Wentao Liu, Baorong Zhou, Mingyu Ou

et al.

Frontiers in Energy Research, Journal Year: 2024, Volume and Issue: 11

Published: Jan. 10, 2024

With the advent of energy Internet and swift growth unified systems, comprehensive demand users has gradually become a problem that cannot be ignored for planning integrated systems. Aiming at this problem, paper suggests multi-agent approach electricity gas, considering users’ holistic consumption behavior. First, utilizing combined subjective objective weighting method, study establishes utility model characteristics. The analysis behavior is conducted through an evolutionary game. On basis, revenue grid gas network investors formulated, game mechanism different analyzed. A dynamic electricity–gas proposed. Ultimately, resolved using iterative exploration approach. validity efficacy proposed method are confirmed simulation example.

Language: Английский

Citations

3
Demand Forecasting and Allocation Optimization of Green Power Grid Supply Chain Based on Machine Learning Algorithm: A Study Based on the Whole-Process Data of Power Grid Materials DOI Open Access

Hanyu Rao,

Jiancheng Li,

Xiaojun Sun

et al.

Sustainability, Journal Year: 2025, Volume and Issue: 17(3), P. 1247 - 1247

Published: Feb. 4, 2025

The efficient management of the green power grid supply chain is great significance in addressing global energy transformation and achieving sustainable development goals. However, traditional methods struggle to effectively cope with complexity dynamics demand forecasting multi-objective optimization problems material allocation. In response this challenge, paper proposes a machine-learning-based allocation method, aiming improve efficiency reduce environmental impacts. First, based on whole-process data materials, multi-model fusion strategy adopted for forecasting. By combining machine learning models such as long short-term memory (LSTM), extreme gradient boosting (XGBoost), random forest, prediction accuracy generalization ability model are significantly improved. Moreover, “distributed collaborative algorithm” proposed. decomposing regions, optimizes transportation routes inventory management, comprehensively reduces transportation, inventory, protection costs while taking into account real-time requirements complex environment. Finally, an empirical analysis carried out combination optimized plan, verifying practical effectiveness proposed method. results indicate that scheme outperforms method terms total cost, efficiency, carbon emissions. Specifically, achieves 13% reduction costs, 10% decrease 25% cut expenses. Additionally, it decreases transportation-related emissions by approximately 250 tons. has obvious economic advantages chain. integrating various algorithms, enhances stability substantially reducing operating This line goals development. provides framework value managing industry.

Language: Английский

Citations

0
Data-Driven Decision-Making for SCUC: An Improved Deep Learning Approach Based on Sample Coding and Seq2Seq Technique DOI Creative Commons
Nan Yang,

Juncong Hao,

Zhengmao Li

et al.

Protection and Control of Modern Power Systems, Journal Year: 2025, Volume and Issue: 10(2), P. 13 - 24

Published: March 1, 2025

The electricity industry has witnessed increasing challenges in power system operation and rapid developments of artificial intelligence technologies the last decades. In this context, studying approach security-constrained unit commitment (SCUC) decision-making with high adaptability precision is great importance. This paper proposes an improved data-driven deep learning (DL) approach, following sample coding Sequence to (Seq2Seq) technique. First, encoding decoding strategy utilized for high-dimensional matrix dimension compression. A DL SCUC decision model based on a Seq2Seq network gated recurrent units as neurons then constructed, mapping between load on/off scheme established through massive data from historical scheduling. Numerical simulation results IEEE 118-bus test demonstrate correctness effectiveness proposed approach.

Language: Английский

Citations

0
TMAC: a Transformer-based partially observable multi-agent communication method DOI Creative Commons
Xuesi Li, Shuai Xue, Ziming He

et al.

PeerJ Computer Science, Journal Year: 2025, Volume and Issue: 11, P. e2758 - e2758

Published: April 4, 2025

Effective communication plays a crucial role in coordinating the actions of multiple agents. Within realm multi-agent reinforcement learning, agents have ability to share information with one another through channels, leading enhanced learning outcomes and successful goal attainment. Agents are limited by their observations ranges due increasingly complex location arrangements, making collaboration based on difficult. In this article, for some partially observable scenarios, we propose Transformer-based Partially Observable Multi-Agent Communication algorithm (TMAC), which improves extracting features generating output messages. Meanwhile, self-message fusing module is proposed obtain from sources. Therefore, can achieve better communication. At same time, performed experimental verification surviving StarCraft Challenge (SMAC) environments where had local observation could only communicate neighboring two test environments, our method achieves an improvement performance 6% 10% over baseline algorithm, respectively. Our code available at https://gitee.com/xs-lion/tmac .

Language: Английский

Citations

0
An intelligent quality prediction and autonomous decision system for natural products manufacturing processes DOI
Qilong Xue, Yang Yu,

Shixin Cen

et al.

Computers & Industrial Engineering, Journal Year: 2024, Volume and Issue: 191, P. 110143 - 110143

Published: April 10, 2024

Language: Английский

Citations

2
Multi-objective explainable smart dispatch for integrated energy system based on an explainable MO-RL method DOI
Jiaming Dou, Xiaojun Wang, Zhao Liu

et al.

Computers & Electrical Engineering, Journal Year: 2024, Volume and Issue: 118, P. 109417 - 109417

Published: July 1, 2024

Language: Английский

Citations

2
Parameter identification method of load modeling based on improved dung beetle optimizer algorithm DOI Creative Commons
Chao Xing, Xinze Xi, Xin He

et al.

Frontiers in Energy Research, Journal Year: 2024, Volume and Issue: 12

Published: Aug. 20, 2024

The role of load modeling in power systems is crucial for both operational and regulatory considerations. It essential to develop an effective reliable method optimizing parameter identification. In this paper, the dung beetle algorithm improved by using good point set, a model identification strategy based on set optimization (GDBO) within framework measurement-based method. proposed involves utilizing PMU voltage data as input, selecting comprehensive model, refining initialization process mitigate influence local maxima. Through iterative objective function Dung Beetle Optimizer (DBO) algorithm, optimal parameters are determined, enhancing model’s ability accurately capture curve. Analysis examples pertaining PMU-measured reveals that GDBO which incorporates outperforms alternative methods such differential evolution (IDE), particle swarm (PSO), grey wolf (GWO), conventional DBO algorithm. This demonstrates superior performance introduced approach context

Language: Английский

Citations

2