Details in GA code

Code details in GA-BLX crossover-gaussion variations:

I wrote a GA code, and on the 2005 test function, I was never able to get results of an order of magnitude similar to that of other students in the lab. Now, by referring to the code of other students, we can find at least the following problems:

1. After crossover and mutation operations, each dimension of the new individual should have a lower limit;

2. Perform a crossover operation on two individuals at random. If the two individuals are the same, no real-value crossover is performed. The previous Code ignores this issue. In the newly modified code, if the two individuals are the same, the next individual is selected for crossover;

3. the results may not be better after the mutation operation. First, stay with the elite individual nremain, and then select N-nremain for cross operations.

Details in GA code