The main problems in the application of any election advertisement are:
1。 Select an appropriate number of representatives.
2。 There are sufficient evaluation functions for the estimated "fitness-producing individuals"
In the search space (point).
In the simplest representation, each gene represents the dimension of the search space. Each dimension can represent an appropriate set of values, discrete or continuous, and a function can. In a simple case, the use of these statements means the existence of binary genes.
Or functions that are not available. In these views, everyone includes a fixed-length binary string of genes or genotypes, representing a subset of a specific function set. Generally, in complex engineering applications, the use of multivalued attributes is more natural [3].
The second important step in the successful application of election advertisements is to select an appropriate fitness evaluation function in engineering design. The evaluation function provides feedback on everyone's fitness service level. Election advertisements use this feedback to bias the search process to improve the average health level of the population.
Especially the specific problems of fitness function.
Of course, the details of one of the two most commonly used mechanisms include reproductive mutation and recombination (or crossover). Mutation model Single parents reproduce and manage their parents through cloning, and then provide some changes of one or more genes in the offspring genome through modification. Mutation is defined by the mutation operator model, which acts on the genome of future generations by modifying genes and changing the mutation rate, and specifies the number, that is, how many genes will be modified on average. In the recombination mode, parents and children bred by many parents clone and recombine their genomes to produce offspring (numbers).
Traditionally, the linear genome of a fixed-length recombinant is replicated and recombined in the form of linear subdivision marked by a simple crossover operator at the intersection [4].
* * * Evolution process of engineering design related to sharing its attributes.
The process includes [5]:
? A search environment with little prior knowledge.
? Because of the excellent search function, the design of search space is effective for sampling.
? In order to avoid the ability of local optimization
? Able to handle high-dimensional data
? The problem of robustness covers a wide range.
? Provide a variety of feasible solutions
? The region in the world that is most capable of finding solutions.