kevinsieverding/t3m405071-seminararbeit

T3M405071 Semniararbeit

Titel: Optimierung von Lieferantenwahl zur Maximierung von Resilienz in Lieferketten

Notizen

• Zielfunktion

• Encodierung der Daten

  • n-Vektor mit rationalen zahlen
  • optimierungsfunktion polynom n-ter ordnung
  • Ziel ist es die passenden gewichte zu finden

• Optimierungsverfahren

  • Lokale Suche
    • Deterministisch
      • Hill Climbing
      • Gradient Descent
  • Probabilistisch
    • Iterated Local Search
  • Globale Suche
    • Deterministisch
      • Branch and Bound
      • Linear Optimization
    • Probabilistisch
      • Genetische Algorithmen

• Lokale Optimierung

Lieferant A: Preis bis 1000 Stück: 10€ Preis von 1000 Stück bis 2000 Stück: 9€ Preis ab 2000 Stück: 8€ Maximale Kapazität: 4000 Stück Beziehungsindex: 0 - 1 Risiko:

• Qualitätsrisiko
• Mengenrisiko
• Terminrisiko

Ziel: Minimierung der Kosten unter Berücksichtigung von Beziehungs- und Qualitätsindex

selection = []                  # list of selected suppliers
x = []                          # list of ordered quantities per supplier
sum = 0                         # sum of ordered quantities      
inc = X / a                     # increment for each iteration
best = 0                        # best value so far

function addToSupplier(inc):
  i = random(1, n)              # pick random supplier
  if S_i not in selection:      # add supplier to current selection if not already present
    add S_i to selection
  if x_i + inc > M_i:           # check if increment would exceed capacity
    inc = inc - (M_i - x_i)     # reduce increment by delta
    x_i = M_i                   # set capacity to maximum
    addToSupplier(inc)          # add remaining increment to another supplier
  else:
    x_i = x_i + inc             # add increment to supplier

while (sum < X):
  addToSupplier(inc)
  current = Z(selection)
  if (current < best):
    reset(selection, x)
  else:
    best = current
    sum = sum + inc

Features

• Pre-configured to comply with the design guidelines for academic papers in the Wirtschaftsinformatik course of study at DHBW CAS.

• GitHub Template Repository to easily jumpstart a new document.

• GitHub Actions Workflow to build the document and attach the PDF as a realease to the repository when new tags are pushed.

  Example:

  # do some changes...
  git commit          # commit your changes
  git tag 1.0         # set a tag
  git push            # push your changes
  git push origin 1.0 # push the tag

  When the tag is pushed a new workflow is triggered that will build the document and create a new release with the PDF as an attached asset using the tag that you have set.

• The LaTeX package minted for beautifully colored code listings.

  • python and the package Pygments are required.
  • The PDF needs to be build with the --shell-escape flag.

  Check the package documentation for further details.