Exercise 5: Implementation of a Simple Evolution Strategy
Date: Summer Term 2006
Now, its time to implement a simple evolution strategy by yourself. Most of the parts are already
prepared for you as you can see from the documentationhttp://www-md.e-technik.uni-rostock.de First of all, you are strongly advised to read the API documentation.
Unit 1: Download the prepared material and read the API documentation.
Unit 2: Implement themainfunction of the simpleea-program. The fitness function should be simply the sum of the squared xi, which is also known as the sphere model.
Where does this function have its minimum? Test the program by running it for a few generations in order to see whether you did everything correctly or not.
Unit 3: To get a deeper understanding of the nature of an evolutionary algorithm, you should now implement themutation object()-routine by yourself. Also, implement a simple uniform crossover operator that may work on two successive offspring.
Unit 4: You are now almost ready to solve the minimization of the power distribution net- work from exercise 2. Replace the fitness function (i.e., the sphere model as noted above), by the calculation of the total network length (as described in exercise 2).
Questions: How many parametersxido you need? How can you map those parame- ters onto the specification of the distributersDi?
Test your program by doing some runs and report your minimal network length to the teacher.
Have fun, Hagen and Ralf.
1
