I have three RSM equations of defect , thrust , torque for drilling operation.
These three objective i want to minimise. So what is the method for changing three objective into the single objective. So i getting a single regression model.
In RSM, Desirabilty analysis is used for multi-objective optimization.
But as far as non-traditional optimization algorithms are concerned, combine all the equations (defect, thrust and torque) into a single objective equation by considering weightage. For example 0.5*defect+0.3*thrust+0.2*torque (so that 100% is achieved).
But for obtaining better results, you should consider constraints apart from combined objective functions.
You may use Grey relational analysis, GRA, to combine three objectives and convert it into single objective. You have to decide on the weightage of each of the objectives based on their importance.
Find the attachment 1, which has the combined objective function inorder to acumulate the various regression model to get / solve by single run through genetic algorithm.
Attachment 2, reveals the desirability function approach to get or convert the two regression model into single one.
Dr. Senthil suggestion revealed in attachment 1. You can even try with Nilesh suggestion i.e. Grey relation analysis. Lin paper titled "Use of the Taguchi Method and Grey Relational Analysis to Optimize Turning Operations with Multiple Performance Characteristics" will assist you to do GRA.
Thankyou munish gupta sir and Muthukrishnan sir for your valuable suggestions , above attached paper give idea about multi objective genetic optimization. Thanks for the suggestions.
You develop three individual regression equation first. after that, u combine them by weighted summation method. it is basically multiplying the each equation by their suitable weight value and then adding them into single equation.