Nonlinear inversion of potential-field data using a hybrid-encoding genetic algorithm

The genetic algorithm is of advantages to solve an inversion of complex non-linear geophysical equations. Its multi-point searching is able to find the globally optimal solution and avoid falling into a local extremum. The searching efficiency of the genetic algorithm is a key to successfully resolve a geophysical inversion problem in a huge model space with multi-parameters. Encoding mechanism impacts mostly in the searching stage of the genetic algorithm. It sometimes is difficult for a standard genetic algorithm (SGA) to make searching successfully, because the crossover and mutation do not receive most effectively searching in mechanisms with only the binary or decimal encoding. For the binary encoding mechanism the operation of the crossover may produce more new individuals. The decimal encoding mechanism, on the other hand, makes the operation of the mutation searching in a larger range. This paper discusses searching potentials between operators in the binary and decimal encoding and presents a hybrid-encoding genetic algorithm (HEGA) mechanism. The method is based on a hybrid encoding in genetic procedure. The mutation operation is executed with the decimal code and other operations with the binary code. The HEGA guarantees the mutation processing with a high probability. HEGA is beneficial to solving the inversion of complex nonlinear geophysical equations. Synthetic and real-world examples demonstrated advantages of using HEGA in inversion of potential-field data.