GA-AM: Genetic Algorithm with Aggressive Mutation

In this section we present the implementation of GA-MA for solving RCPSP. The executable and input data will be provided to solve the PSPLIB. The proposed algorithm was implemented in c++ and compiled on Visual Studio 2013.

Results:

In the next table we show the deviation (%) from the optimal solution or from the critical path depending on the set.

The makespan obtained by using GA-AM for each problem can download here.

Minimum requirements:

• Windows 10 , Windows 7 Service Pack 1, Windows 8, Windows 8.1. On 64 bits.
• 500mb ram or higher.
• Visual C++ Redistributable Packages for Visual Studio 2015. You can download it here.

Downloadable content:

• Data sets and GA-AM.

Instructions:

1. Decompress the GA-AM.rar in your hard disk. You should have the following directory structure:

   

 

2. GA-AM folder contain the algorithm (0Algorithm_v1.3) and the data input (Arhcivos30, ..., LB0normales).

 

3. In Release folder there is a txt file (datos.txt). It is the file that contains the problems set will be solved.

   

   In this example there will be 4 runs, all over j30 set. The first two with 1000 iterations, next with 5000 iterations and 50000 iterations. For adding more problems sets, add to this file a new line with the format:

   1 "j" "itera" "cores" no no 0 0.05 1.0 0.5 0 0

   Where j can be 30, 60, 90 and 120 for j30, j60, j90 and j120 problem sets from PSPLIB, itera are the iterations and cores are the threads to use. The others are control parameters.

 

4. To run the algorithm, look into Release folder the file: ga-amV1.3.exe, and execute it.

 

5. The sumary of result will be shown in Windows console and will save into 0Algorithm_v1.3 folder, in output.txt file.

 

6. Moreover, each run creates a txt file (for example 1_1000vectorSolSecuencias.txt), identified with a run number plus iterations. Within this file there are the detailed results of each problem of each set.