Research Of Application Of GEP In Wheat Aphid Population Modelling

Wheat is one of the most important crops in our country, which play a significant role in the daily life of citizens and national economy. Wheat aphid is a kind of pest which frequently breaks out in each wheat producing area, not only nibble the nutrition of wheat, but also influence the photosynthesis, could transmit the wheat viruses as well,which lead the decreasing of wheat output and quality, severely threaten the producing of wheat. This paper focus on simulating the relationship between climate factors and wheat aphid with computer technology, but previous mathematical modelling methods have some limitations, so this paper make use of GEP and research data of wheat aphid to model the population of wheat aphid. However, GEP still could not overcome the common problem------stagnation and premature convergence which is always occur in intelligence algorithms, so this paper complete the modelling of wheat aphid population on the basis of improving in terms of this problem.The main tasks of this paper: 3 kinds of improved algorithm on the basis of standard GEP, conduct the simulation experiments with 3 kinds of improved algorithms,which show the performance of them is better than standard GEP and same kind of algorithm in different degree. At last take advantage of 3 algorithms to develop the wheat aphid population modelling software. The concrete contents are:Integrate the standard GEP and frog leaping algorithm on the basis of individual diversity, sort individuals into different frog group in according to the individual diversity, each group evolve alone, which make the individuals which have difference could promote each other, we could see that the thought of frog leaping improve the performance of GEP in certain extension.Adopt the entropy to evaluate the diversity of population, then use adaptive change of 2 kinds of mutation rates on the basis of diversity entropy, additionally remold the crossover and mutation operator, the simulation experiment show that these actions could make the performance of GEP increase.Admired by the thought of fuzzy system clustering, replace the individual whose distance is too near to other individual randomly, which make the search uniformly, and the threshold of close degree could adaptive change with the carrying out of GEP, which is consistent with the trend of algorithm, simulation experiment show the performance of this algorithm have competitiveness.Combine one of above 3 algorithms whose performance is best in the simulation experiment and make use of wheat aphid data to model the relationship of wheat aphid and climate factors to testify the effectiveness of these algorithms in modelling the climate response model of wheat aphid.Develop the wheat aphid modelling software with 3 algorithms, this system includes 3 parts: parameters setting module, backstage computation module, result visualization module, the running operation and testing show that the system has good performance and user experience.