Inventory Simulation ⇄ Optimization
Inputs
Simulation & Optimization controls
Simulation Inputs
Fixed defaults:
Review interval: 1 period
Demand: Poisson arrivals × discrete size
Lead time: exponential
Cost parameters & policy types: Project details, Per-product constants
Variables:
Product A
Product B
Product C
Product D
Product E
Replicates
5
Averages multiple stochastic runs (1–20). Default: 3.
Horizon (periods)
60
Number of review periods (10–1000). Default: 60.
Optimization (GA) Inputs
Fixed defaults:
Time limit: 20 s; ε: 0.0005
Patience: 10 gens; Min gens: 25
Mutation step: 4; Tournament k: 2
Elitism: 2; Random immigrants: 0.15
Adaptive mutation: on
Restart patience: 12
Variables:
Population size
40
Mutation rate
0.30
Crossover rate
0.90
Run Optimization
Set sliders & products in
Inputs
, then press
Run
.
jump to Inputs
Reset to default
Optimization time:
0
ms
Project details
Simulation & Optimization
Inventory Simulation
Multi-product, continuous-review inventory model.
Policies:
(r,Q)
(fixed lot) or
(s,S)
(order-up-to).
Random demand (Poisson arrivals × demand size) and random lead times (exponential).
Tracks holding, shortage, and ordering costs plus service levels.
Horizon of 60 review periods; results averaged over replicates.
Per-product constants
Product A
(s,S)
· S=50
· init=50 · backlog=yes
λ (arrivals/period): 10
Lead time mean: 3
Holding h: 1
Shortage p: 5
Setup K: 32
Unit c: 3
Sizes = [0.5, 1, 1.5], Probs = [0.33, 0.33, 0.33]
Product B
(r,Q)
· Q=15
· init=30 · backlog=no
λ (arrivals/period): 8
Lead time mean: 2
Holding h: 1
Shortage p: 5
Setup K: 32
Unit c: 3
Sizes = [1, 2, 3], Probs = [0.20, 0.50, 0.30]
Product C
(s,S)
· S=40
· init=40 · backlog=yes
λ (arrivals/period): 6
Lead time mean: 4
Holding h: 0.8
Shortage p: 6
Setup K: 28
Unit c: 2.5
Sizes = [1, 1.5, 2], Probs = [0.40, 0.40, 0.20]
Product D
(r,Q)
· Q=20
· init=25 · backlog=no
λ (arrivals/period): 12
Lead time mean: 2.5
Holding h: 1.2
Shortage p: 5
Setup K: 35
Unit c: 2.8
Sizes = [0.5, 1, 2], Probs = [0.30, 0.50, 0.20]
Product E
(s,S)
· S=30
· init=30 · backlog=yes
λ (arrivals/period): 5
Lead time mean: 3.5
Holding h: 0.9
Shortage p: 7
Setup K: 30
Unit c: 2.7
Sizes = [0.5, 1, 1.5], Probs = [0.20, 0.60, 0.20]
Optimization (Genetic Algorithm)
Searches for good reorder points across products.
Population-based: selection, crossover, and mutation evolve solutions.
Objective: minimize average cost per period across replicates.
Stops when time budget expires or cost improvement falls below ε (patience).
History includes best/mean costs and per-product metrics.
References
Holden, L. (2017).
Inventory Optimization Using a SimPy Simulation Model
.
Strijbosch, L.W.G. & Moors, J.J.A. (2002).
Simulating an (R, s, S) Inventory System
.
Pasandideh, S.H.R., et al. (2011).
Genetic algorithm for multi-product EOQ inventory control
.
Maitra, S. (2024).
Inventory Management Under Stochastic Demand: Simulation-Optimization
.