HAL Id: hal-01240450
https://hal.archives-ouvertes.fr/hal-01240450
Submitted on 9 Dec 2015
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Distributed under a Creative CommonsAttribution - NonCommercial - ShareAlike| 4.0 International License
Strategic and Tactical Urban Farm Design
Nicolas Brulard, Van-Dat Cung, Nicolas Catusse
To cite this version:
Nicolas Brulard, Van-Dat Cung, Nicolas Catusse. Strategic and Tactical Urban Farm Design. FSD5 5th International Symposium for Farming Systems Design, Sep 2015, Montpellier, France. �hal- 01240450�
Graphical representation of the results
STRATEGIC AND TACTICAL URBAN FARM DESIGN
1 Univ. Grenoble Alpes, G-SCOP, F-38000 Grenoble, France CNRS, G-SCOP, F-38000 Grenoble, France
2 Jardins de Gally, 78870 Bailly, France
nicolas.brulard@grenoble-inp.org
Nicolas Brulard 1,2 , Van-Dat Cung 1 , Nicolas Catusse 1
• Model to support decision in creating new urban farms and adapting existing peri-urban farms with viable business models
• Multi-techniques, multi-products and multi-clients systems producing perishable products on limited surfaces with expensive labor cost
• Considering fresh products perishability requires to include tactical scheduling decisions in our strategic sizing problem
Context and problem description
Data instances & computation times MILP model structure
Prospects
• Short-time perishability (some days)
• Response to daily demands/prices
On-plot and cold room limited storages, loss functions, price reduction
Considering products perishability
Different management practices can be needed to match these demands
in fresh products
Objective :
Preprocessing and optimization
Sales
Amortization Tasks cost Labor cost
• 5 to 9 sets of management practices, 1 or 2 products, 4 to 9 plots max, 1 or 2 clients.
• Optimal solution reached in 8 to 1100s (Cplex 12.6, Intel Core i7- 4600U 2.10GHz proc, 8.0 Go of RAM).
Model tested on real farm data (lettuce, strawberry and tomato farm)
New formulation of lot sizing constraints and better optimization algorithms to speed up resolution
Integration of multi-year strategic sizing, robust solutions.
-1j << >> 3j 0 0 0
MP1 MP2 MP3 MP4 MP5 MP6 MP7 MP8 MP9
Total area affectation
0 m² 0 m² 0 m²
3000 m²
511 m² 161 m²
0 m²
1059 m²
667 m²
0 1000 2000 3000 4000
Surface (m²)
Production of each set of management practices
Cumulated production and stocks
Required labor
Demand 𝑐,𝑝 ∆𝑐,𝑝,𝑑
On-plot stocks and picked quantities
Product 1 cold stock (Sets of MP 1 to 5)
Product 2 cold stock (Sets of MP 6 to 9)
𝑥𝐷𝐶𝑐,𝑝,𝑑𝜃,𝛽 𝑥𝐷𝐶𝐿𝜃,𝛽𝑐,𝑝,𝑑
𝑥𝐶𝐿𝜃,𝛽𝑝,𝑑 𝑥𝐶𝑝,𝑑𝜃,𝛽
Hired labor 𝐿𝑇𝑑
Tasks time Picking time 𝐿𝑇𝑥𝑃𝑖,𝑑𝜃
Monitoring time
max 𝑅 =
𝑝,𝜃,𝛽,𝑑
𝑥𝐷𝐶
𝑐,𝑝,𝑑𝜃,𝛽∗ 𝑃𝑅
𝑐,𝑝,𝑑∗ (1 − 𝜔
𝑖,𝑝𝜃,𝛽)
−
𝑖
(𝐴𝑚𝑜𝑓
𝑖+ 𝑎
𝑖∗ 𝐴𝑚𝑜𝑣
𝑖) −
𝑖,𝑡
𝐴
𝑖∗ 𝐶
𝑖𝑡−
𝑑
𝐿𝑇
𝐷∗ 𝐶𝐿𝑇 +
𝑑
𝐶𝐿𝑇
±∗ (𝐿𝑇
𝑑++ 𝐿𝑇
𝑑−)
Production curve 𝑌𝑖,𝑑
+ 2-period shift
- 2-period shift Av. on-plot stock 𝑥𝑖,𝑑𝜃
Picked quantities 𝑥𝑃𝑖,𝑑𝜃 MP1
MP2 MP3 MP4
MP5
MP6
MP7 MP8
MP9 PARAMETERS
VARIABLES
PREPROCESSING
OPTIMIZATION
Set of management practices i
𝑌
𝑖,𝑑0 25 50
1 2 3 4 5 6 7 8 9 10
0 25 50
1 2 3 4 5 6 7 8 9 10
0 25 50
1 2 3 4 5 6 7 8 9 10
0 25 50
1 2 3 4 5 6 7 8 9 10
0 1
ta1ta2
0 1
ta1ta2
0 1
ta1ta2
0 1
ta1ta2
0 1
ta1ta2 0 0.5 1
1 2 3 4 5 6 7 8 9 10
𝑌
𝑖,𝑑ta1 ta2
𝐴𝑃
𝑖,𝑑 ,𝑡𝐷
𝑖𝑡𝐷𝐷
𝑖𝑡(kg/m²) Production curve
(periods) Tasks starting date
(periods) Tasks starting date
(kg/m²) Shifted prod curve
(periods) Shifted task avail.
Unauthorized shift, starting date lower than .𝐷𝑎 𝑖
𝑎
,𝑑 ,𝑡𝑎
,𝑑 ,𝑡𝑎
,𝑑 ,𝑡𝑎
,𝑑 ,𝑡𝑎
,𝑑 ,𝑡(m²)
𝐴
𝑖= 𝑎
𝑖=
𝑖,𝑑 ,𝑡,𝑑
𝑥
𝑖,𝑑𝜃𝑥 , 𝑥 , 𝑥 ,
∆
𝑐,𝑝,𝑑0 2 4 6 8 10
1 2 3 4 5 6 7 8 9 10
𝐿𝑇𝑑 = 𝐿𝑇𝑥𝐷𝐶𝑝,𝑑𝜃,𝛽
𝑝,𝜃,𝛽 + 𝐿𝑇𝑥𝑃𝑖,𝑑𝜃
𝑖,𝜃 + 𝐿𝑇 𝑖,𝑑
𝑖 (m²/period)
Av. production (kg)
𝑥𝑃
𝑖,𝑑𝜃Picked quantity (kg)
𝑥𝐶
𝑝,𝑑𝜃,𝛽Avail. cold room stock (kg)
𝑥𝐷𝐶
𝑐,𝑝,𝑑𝜃,𝛽Delivered quantity (kg)
Sets / products matrix
𝐿𝑇
𝑑Hired labor time (h/period)
Demand
Other sets of MPs picked quantities 𝑝,𝑖