How to Normalize Objective Function in HyperStudy ?

BUGRA
BUGRA Altair Community Member
edited March 2023 in Community Q&A

Hello, 

I am trying to conduct Multi-Objective Optimization  Analysis. Mainly, my design variables are  thickness of rib and skin panel, 

Response I am looking for is Frequency and element stress.  I have constraints on them.

Objectives are to minimize tip displacement and mass. When I selected weighted sum method, it only multiplies weights with tip displacement and mass. Then, sum them up . I would like to normalize both of them. Then, multiply them with weight coefficients. I can't find appropriate selection in tool bar in hyperstudy. 

How can I activate normalizing objective functions, 

Thanks

1.JPG 28.6K
Tagged:

Answers

  • Adriano_Koga
    Adriano_Koga
    Altair Employee
    edited February 2023

    you could simply build an extra response 'myWeightedObj' and create an expression with some weighting factors as desired.

     

    let's say you have

    r_1 = disp

    r_2 = stress

    r_3 = mass

     

    myWeightedObj would be, i.e:

    (1/10)*r_1 + (1/100)*r_2 + 1*r_3

     

    Of course you would need to tweek a littkle bit the weightings.

  • BUGRA
    BUGRA Altair Community Member
    edited March 2023
    Adriano A. Koga_21884 said:

    you could simply build an extra response 'myWeightedObj' and create an expression with some weighting factors as desired.

     

    let's say you have

    r_1 = disp

    r_2 = stress

    r_3 = mass

     

    myWeightedObj would be, i.e:

    (1/10)*r_1 + (1/100)*r_2 + 1*r_3

     

    Of course you would need to tweek a littkle bit the weightings.

    Hello Adriano, 

    Thanks for responding, 

    I see you are directly multiplying objective functions with coefficient, how can I normalize them ?

Categories