liu.seSearch for publications in DiVA
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Energy Management Based on Neural Networks for a Hydraulic Hybrid Wheel Loader
Linköping University, Department of Management and Engineering, Fluid and Mechatronic Systems. Linköping University, Faculty of Science & Engineering.ORCID iD: 0000-0002-2781-3752
Linköping University, Department of Management and Engineering, Fluid and Mechatronic Systems. Linköping University, Faculty of Science & Engineering.ORCID iD: 0000-0002-3877-8147
Volvo Construction Equipment.
Volvo Construction Equipment.
Show others and affiliations
2020 (English)In: 2020 IEEE Global Fluid Power Society PhD Symposium, 2020Conference paper, Published paper (Refereed)
Abstract [en]

This paper presents a method to derive optimized energy management strategies for a hydraulic hybrid wheel loader. Energy efficiency is a key aspect for the sustainability of off-road mobile machines. Energy management strategies for on-road hybrid vehicles cannot be directly applied to off-road hybrid machines. One significant reason is that there are added degrees of freedom with respect to how power can be recovered, exchanged and reused in the different functions, such as drivetrain or work functions. This results in more complex energy management strategies being derived. This paper presents an analysis and preliminary conclusions for a proposed method to derive optimised online energy management strategies for a hydraulic hybrid wheel loader. Dynamic programming is used to obtain optimal offline energy management strategies for a series of drive cycles. The results are used as examples to train a neural network. The trained neural network then implements the energy management strategy and is used to make optimised control decisions. Through simulation, the neural network’s ability to learn the dynamic programming decision-making process is shown, resulting in the machine operating with fuel consumption similar to that of the offline optimal energy management strategy. Aspects of simplicity to model these machines for dynamic programming optimisation, the data necessary to train the network, the training process, variables used to learn the dynamic programming decision-making process and the robustness of the network when facing unseen operational conditions are discussed. The paper demonstrates the simplicity of the method for taking into account variables that affect the control decisions, therefore achieving optimised solutions.

Place, publisher, year, edition, pages
2020.
Keywords [en]
Construction Machines, Hydraulic Hybrid, Energy Management Strategies
National Category
Energy Systems
Identifiers
URN: urn:nbn:se:liu:diva-173170ISBN: 9781728141381 (electronic)OAI: oai:DiVA.org:liu-173170DiVA, id: diva2:1526470
Conference
2020 IEEE Global Fluid Power Society PhD Symposium, Guilin China, October 19-21, 2020
Available from: 2021-02-08 Created: 2021-02-08 Last updated: 2022-08-16Bibliographically approved
In thesis
1. On Machine Learning-Based Control for Energy Management in Construction Machines
Open this publication in new window or tab >>On Machine Learning-Based Control for Energy Management in Construction Machines
2022 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

High energy efficiency is a key requirement for modern construction machinery. This is because of stricter environmental targets, electrification, and reduction of operation costs. To meet this requirement, the powertrain architectures of the machines are becoming increasingly complex, for example through hybridisation of drivetrain and work functions, or with improved hydraulic systems. However, the more complex the architecture is, the harder the management of splitting power between different sources and consumers. The number of work functions, operating environments, and tasks these machines engage in, along with the added degrees of freedom with respect to how energy can be recovered, ex-changed, and reused, makes them unique. Therefore, the development of control strategies for energy management in such machines requires specific research and development with their architecture and application in focus. This doctoral thesis presents an analysis of two methods for the development of machine learning-based energy management strategies for construction machines. One is based on supervised learning and the other on reinforcement learning. The methods use optimisation to find optimised solutions for the control problem of the systems and machine learning for learning and implementing the control decisions. In both methods, models of the physical systems are used for the learning and training. The thesis highlights and confirms, with experimental results, the potential of such methods to derive control strategies for these machines. The studied methods can learn and implement improved control decisions in the real systems that result in the potential for increased efficiency. At the same time, their robustness is shown in the application to unseen scenarios during training, although that does not eliminate the need for further training in the real systems after deployment. The thesis also increases the comprehensiveness on energy management for construction machines. The thesis was completed in a double-degree format between the Federal University of Santa Catarina, Florianópolis, Brazil, and Linköping University, Linköping, Sweden.

Abstract [sv]

Hög effektivitet är ett nyckelkrav för moderna byggmaskiner. Detta på grund av strängare miljömål, elektrifiering och sänkta driftskostnader. För att möta detta krav blir maskinernas arkitektur mer komplex, till exempel genom hybridisering av drivlina och arbetsfunktioner. Men ju mer komplex arkitekturen är desto svårare blir hanteringen av maktdelning mellan olika källor och konsumenter. Antalet arbetsfunktioner, drifts-miljö och uppgifter de engagerar tillsammans med de ökade frihetsgraderna med avseende på hur kraft kan återvinnas, bytas ut och återanvändas, gör dem unika. Därför kräver utvecklingen av styrstrategier specifik utveckling med deras arkitektur och tillämpning i fokus. Denna doktors-avhandling presenterar en analys av två metoder för utveckling av optimerade och intelligenta energihanteringsstrategier i realtid för delsystem av komplexa entreprenadmaskiner. De utvärderade metoderna använder optimering för att hitta optimala lösningar för systemens kontrollproblem, och använder maskininlärning som ett sätt att lära sig och implementera de optimerade besluten. I båda metoderna används modeller för lärandet och träningen. Avhandlingen belyser och bekräftar experimentellt potentialen hos sådana metoder för att härleda kontrollstrategier för dessa maskiner. De studerade metoderna kan lära sig och implementera optimerade styrbeslut i de verkliga systemen vilket leder till ökad effektivitet. Samtidigt visar det sig att de är robusta mot osynliga scenarier under träning, även om det inte eliminerar behovet av vidare-utbildning i de verkliga systemen efter utplacering. Examensarbetet ökar också heltäckningen om energihantering för entreprenadmaskiner. Avhandlingen har utvecklats i ett dubbelgradersformat med Federal University of Santa Catarina, Florianópolis, Brasilien och Linköpings Universitet, Linköping, Sverige.

Abstract [pt]

Alta eficiência energética é um requisito para máquinas de construção modernas, sendo este uma consequência das metas ambientais, eletrificação e redução de custos. Para atender este requisito as arquiteturas dos trens de potência das máquinas têm se tornado mais complexas, por exemplo, através da hibridização do sistema de tração e funções de trabalho, ou  sistemas hidráulicos melhorados. Entretanto, quanto mais complexa a arquitetura, mais difícil se torna o gerenciamento da divisão de potência entre as fontes e consumidores. O número de funções de trabalho, ambientes de operação e tarefas que elas executam, juntamente com os graus de liberdade relacionados à como energia pode ser recuperada, trocada e reutilizada, às tornam únicas. Dessa maneira, o desenvolvimento de estratégias de controle requer pesquisa e desenvolvimento específicos com as suas arquiteturas e aplicações em foco. Esta tese de doutorado apresenta uma análise de dois métodos para o desenvolvimento de estratégias baseadas em aprendizado de máquina para o gerenciamento de energia em máquinas de construção. Um é baseado em aprendizado supervisionado e outro em aprendizado por reforço. Os métodos avaliados usam otimização para encontrar soluções otimizadas para o problema de controle dos sistemas, e usam aprendizado de máquina como meio para aprender e implementar as decisões de controle. Em ambos os métodos, modelos dos sistemas físicos são utilizados para o aprendizado e treinamento. A tese destaca e confirma através de resultados experimentais, o potencial destes métodos em obter estratégias de controle para estas máquinas. Os métodos estudados são capazes de aprender e implementar decisões de controle melhores nos sistemas reais resultando em potencial aumento de eficiência energética. Ao mesmo tempo, é mostrado a sua robustez na prática a cenários não vistos durante o treinamento, apesar de isso não eliminar a necessidade de continuar o treinamento depois de implementadas no sistema real. A tese também aumenta a compressão sobre gerenciamento de energia em máquinas de construção. A tese foi desenvolvida em formato de cotutela com a Universidade Federal de Santa Catarina, Florianópolis, Brasil, e a Universidade de Linköping, Linköping, Suécia.

Place, publisher, year, edition, pages
Linköping: Linköping University Electronic Press, 2022. p. 104
Series
Linköping Studies in Science and Technology. Dissertations, ISSN 0345-7524 ; 2244
Keywords
Machine Learning, Energy Management, Construction Machines, Aprendizado de Máquina, Gerenciamento de Energia, Máquinas de Construção
National Category
Mechanical Engineering
Identifiers
urn:nbn:se:liu:diva-187244 (URN)10.3384/9789179294168 (DOI)9789179294151 (ISBN)9789179294168 (ISBN)
Public defence
2022-09-12, Rum A38, A-Building, Campus Valla, Linköping, 10:15 (English)
Opponent
Supervisors
Available from: 2022-08-16 Created: 2022-08-16 Last updated: 2022-08-29Bibliographically approved

Open Access in DiVA

No full text in DiVA

Authority records

Raduenz, HenriqueEricson, LiselottKrus, Petter

Search in DiVA

By author/editor
Raduenz, HenriqueEricson, LiselottKrus, PetterDe Negri, Victor J.
By organisation
Fluid and Mechatronic SystemsFaculty of Science & Engineering
Energy Systems

Search outside of DiVA

GoogleGoogle Scholar

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 572 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf