• 1 reply
  • Latest Post - ‏2013-03-04T14:49:25Z by GGR
554 Posts

Pinned topic Probability for jobs to move from one machine to the other

‏2013-03-02T17:32:44Z |
I am writing a code in "IBM ILOG CPLEX Optimization Studio" to sequence jobs on 2 machines (to minimize total weighted tardiness). A .mod tab has the code while the .dat tab has the data.

The 2 machine system I am trying to set up is as follows:
1. We need to sequence jobs on the 1st machine.
2. Then (after completing 1st machine), the jobs have a 10% probability of going to the second machine, and a 90% probability of leaving the system (after completing 1st machine).
3. Again, if it goes to the second machine (based on 10% probability), it has to return to the 1st machine.
4. Upon returning to the 1st machine and completing the processing(for a second time), it should leave the system.

Presently, I am coding as below, which means that it will go from machine 1 to 2 (100% prob.), and then back to machine 1 before leaving the system. The tuplet represents the Machine # and the Processing time at that machine.

<0,17>, <1,9>, <0,17> {Machine 0 represents first machine, while 1 represents second machine}.

Is there a way that I can include a probability parameter to code it as required? Please help!!!!

Thanks and Regards,
Updated on 2013-03-04T14:49:25Z at 2013-03-04T14:49:25Z by GGR
  • GGR
    73 Posts

    Re: Probability for jobs to move from one machine to the other


    Yo are mistaken about IBM ILOG CPLEX Optimization Studio. That is a tool that solve combinatorial problems. That is the variable are not stochastic and the constraint make that there is no polynomial algorithm to solve the model.

    You can anyway combine your stochastic model with CPO by generating a set of instances of your stochastic problem. Each instance is a model for CPO that solves the instance.

    In you case, the instances are operations linked by precedence that follows all your possible mode in 1,2,3 and 4 points. The set of operations of the set statistically follows the rules you gave.