Genetic Algorithm for Orderplanning

Requirements

• Python 3.7.3
• PyGAD 2.18.1

Input CSV data examples (Do NOT include headers in file)

• Copy and edit the example CSV files from example input data\ to imports\

Part & operations (The parts that can be produced)

• Filename input_parts.csv
• First column Part name
• Add as many Operations as needed in following columns

Part name	Operation	Operation	Operation	Operation
Part1	Op1	Op2	Op3
Part2	Op1	Op2
Part3	Op1
Part4	Op1	Op2	Op3	Op4
Part5	Op1	Op2	Op3

Machines/Assets (The assets that can produce Parts)

• Filename input_machines.csv
• Part name & Operation name must match a Part in the Parts input
• Time is the time required for the operation in the machine in minutes

Name	Part name	Operation name	Time
Machine1	Part1	Op1	30
Machine1	Part1	Op2	25
Machine1	Part1	Op3	30
Machine1	Part2	Op1	25
Machine1	Part2	Op2	60
Machine1	Part2	Op3	55
Machine2	Part1	Op1	25
Machine2	Part1	Op2	60
Machine2	Part2	Op2	100

Work order list (The list of Parts that will be produced)

• Filename input_workorders.csv
• First column is the Part name to be produced
• Amount of parts to be produced

Part name	Amount
Part1	200
Part2	100
Part2	100
Part3	50
Part1	60
Part4	120
Part4	30
Part1	50
Part5	100

Output schedule

• A schedule result file is output in the format exports\schedule_export_{run_id}_{total_time_in_minutes}.csv

Machine name	Order id	Order size	Part name	Operation	Dependencies	Scheduled	Starttime	Endtime
Machine2	4	60	Part1	Op1		True	0	1500
Machine2	0	200	Part1	Op1		True	1500	6500
Machine2	7	50	Part1	Op1		True	6500	7750
Machine2	setup	setup	setup	setup		True	7750	7810
Machine2	7	50	Part1	Op2	['Op1']	True	7810	10810
Machine2	setup	setup	setup	setup		True	10810	10870
Machine2	2	100	Part2	Op2	['Op1']	True	10870	20870
Machine1	2	100	Part2	Op1		True	0	2500
Machine1	1	100	Part2	Op1		True	2500	5000
Machine1	setup	setup	setup	setup		True	5000	5060
Machine1	0	200	Part1	Op2	['Op1']	True	6500	11500
Machine1	setup	setup	setup	setup		True	11500	11560
Machine1	4	60	Part1	Op3	['Op1', 'Op2']	True	11560	13360
Machine1	7	50	Part1	Op3	['Op1', 'Op2']	True	13360	14860
Machine1	0	200	Part1	Op3	['Op1', 'Op2']	True	14860	20860

GA configuration

The file ga_config.json can be used to edit parameters of the Genetic Algorithm

{
    "ga_parameters" : {

        "population_size" : 50,
        "generations" : 10000,
        "parents_mating" : 10,
        "keep_parents" : 5,
        "mutation_percent_of_genes" : 20,
        "saturation" : 500
    
        }
}

Run the application

To start the application, run the command: python ga.py --{optional arguments}

Arguments

• --no_plot (Do not show the plot, default false)
• --runs {x} (Number of runs(x), creates a separate schedule for every run, default 1)
• --no_csv_export (Do not export a CSV file, default false)

Example: python ga.py --no_plot --runs 3 will create 3 output schedules in the exports folder, without displaying the plot during the runs.