Download presentation
Presentation is loading. Please wait.
Published bySolomon Simpson Modified over 9 years ago
1
Jian Ren, Mark Harman CREST Centre, SSE Group, UCL 23 Feb, 2011, 11 th COW
2
Research Motivations Solutions Representations: WPO and Staffing Fitness Evaluation: Simulation of Execution Cooperative Coevolution Process Empirical Results and Issues 2 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
3
Staff Assignments (Team Construction) How to assign staff into project teams? Work Packages Scheduling (WP Ordering) How to put the WPs in a good order to execute? Objective Find earlier completion time of a project by optimising TC and WPO simultaneously 3 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
4
How to assign N staff into M teams? Staff 1 Staff 2 Staff 3 Staff 4 Staff 5 …Staff 29 Staff 30 To Team 34235…32 N staff Species #1: assignment of staff to teams 4 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
5
5 34235…32 Parent_1
6
34235…32 6 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman 42524…21 42525…32 34234…21 Parent_1 Parent_2 Child_1 Child_2 Single Point Crossover
7
Teams’ Capacity T1T2T3T4T5 4 staff6 staff8 staff5 staff7 staff 7 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman Staff 1 Staff 2 Staff 3 Staff 4 Staff 5 …Staff 29 Staff 30 To Team 34235…32
8
How to put the WPs in a perfect or near perfect order? Perfect ordering = All teams are busy all the time and finish their last WP at the same time Solution: Attributes of WP ID Efforts: total efforts required by executing the WP Dependency: ID of its predecessor WP1WP4WP6…WP56WP60 Species #2: orderings of the WPs 8 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
9
WP1WP4WP6…WP56WP60 9 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman Parent_1
10
WP1WP4WP6…WP56WP60 10 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman Parent_1 WP2WP3WP1…WP60WP59Parent_2 WP1WP4WP6…WP60WP59Child_1 WP2WP3WP1…WP56WP60Child_2 B: the rest of the other parent after WPs in A are deleted A: copied from one parent
11
Objective Earlier Overall Completion Time = Higher Fitness WP1WP4…WP60 Efforts8412 Processing Simulator The `Best` available team picks the next WP before other available teams. First Come First Served One WP blocks the queue of waiting WPs until its predecessors are all finished. 11 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman Staff 1 Staff 2 Staff 3 …Staff n-1 Staff n To Team 342…32
12
Processing Simulator Objective Earlier Overall Completion Time = Higher Fitness T1T2…T5 Capacity6 staff8 staff…7 staff Date (it becomes available) 00…0 WP1WP4…WP60 Efforts8412 Start Time Finish Time To Team 12 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
13
Processing Simulator Objective Earlier Overall Completion Time = Higher Fitness T1T2…T5 Capacity6 staff8 staff…7 staff Date (it becomes available) 01 st Day …0 WP1WP4…WP60 Efforts8412 Start Time 0 th Day Finish Time 1 st Day To Team T2 13 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
14
Processing Simulator Objective Earlier Overall Completion Time = Higher Fitness T1T2…T5 Capacity6 staff8 staff…7 staff Date (it becomes available) 22 nd Day 21 st Day …21 st Day WP1WP4…WP60 Efforts8412 Start Time 0 th Day 20 th Day Finish Time 1 st Day 0.5 th Day 22 nd Day To Team T2T3T1 14 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman The latest finishing time is the overall completion time of a simulation.
15
Initialisation for both species //(S1: TC, S2: WPO) FOR (E = 1 : External_Generations) FOR (I = 1 : Internal_Generations) internal evolutionary progress for Species #1 end FOR FOR (I = 1 : Internal_Generations) internal evolutionary progress for Species #2 end FOR 15 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
16
16 Internal evolutionary process for TC Internal evolutionary process for WPO Evolved TCEvolved WPO
17
X: Internal Generation Number Y: Overall Completion Time Blue line: Represents the flow of Cooperative Co-evolution Process Upper 4 box-plots: Fitness of TC Lower 4 box-plots: Fitness of WPO Each box-plot represents 4 generations of internal evolutionary process. 17
18
18 Internal Generation Num = 10 External Generation Num = 10 X: Generation Num (Time) Y: Project Completion Time
19
Given the same number of evaluation efforts, CCEA finds better solutions than single population EA. Optimisation on WPO contributes more benefit than on TC. 19 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
20
Configurations for CCEAInternal Generation NumExternal Generation Num I1100 II10 III1001 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman 20 * Configuration III can be considered as single-species EA, as the FOR loop has only been executed once. No interactions between two population have chance to happy before the end of the FOR loop. It is an extreme case of CCEA. 3 sets of configurations are applied to 4 real-world projects. Comparing the earliest finishing time found by I, II and III: Configurations’ Achievement #First Place#Second Place#Third Place I611 II260 III017
21
21 Internal Generation Num = 100 External Generation Num = 1 X: Generation Num (Time) Y: Project Completion Time
22
Given the same number of evaluation efforts, CCEA finds better solutions than single population EA. Optimisation on WPO contributes more benefit than on TC. Different configurations of the CCEA produce results with different characteristics. 22 Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman
23
Problem Model Staffing skills, changing teams during the project, absence, etc WP changed during the project, mis-estimate efforts The configurations of CCEA is complex. Diversity of the species of Team Constructions, Processing Simulator (execution of WPs) Job-shop scheduling permutation, etc Comparison and Validation Single-species EA (Statistics Analysis) Verification by More Real-world Data from Software Project Management, Cloud Computing Cooperative Coevolutionary Optimisation on Work Package Scheduling and Staff Assignments Jian Ren & Mark Harman 23
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.