Research And Application Of Moth-flame Optimization Algorithm Based On Chaotic Crisscross Operator

Moth Flame Optimization(MFO)is a new meta heuristic algorit hm proposed by Australian scholar Seyedali Mirjalili in 2015.The algorithm extends the model of the spiral curve movement of moths around flame in nature to an intelligent algorithm which is suitable for function solving and parameter optimization.Although the MFO algorithm is easy to implement and adjust parameters less,and has a good solut ion performance.However,when the MFO algorithm solves the complex function of large scale,it is still prone to fall into the local optimal solution,and the efficiency of the calculation is slow and can not converge quickly.Aiming at these two shortcomings of MFO algorithm,the moth-flame optimization algorithm which is based on chaotic crisscross operator(CCMFO)algorithm is proposed.First,we introduced the crisscross chaotic mechanism in the basic MFO algorithm.Both longitudinal and lateral fla me cross were complementary.When the horizontal dimension artificia l moths falled into the local optimal solution,the local optimal solution could be jumped through the longitudinal cross calculation.When the artificial moth in the longitudinal dimension falled into the local optimal solution,it could be expanded through the flame information brought by the crosswise cross.A better flame solution was avoided to avoid premature algorithm.Two dimensional cross optimization could make flame information spread wider and faster in the population,and improve the accuracy and speed of algorithm.At the same time,through the addition of chaos operator in the crisscross,the relative parameters were calculated through its randomness and ergodicity to control the position change of the artificial moth and avoid the precocity of the algorithm.The crisscross chaotic mechanis m made up for the two shortcomings of the slow convergence speed and low accuracy of the basic MFO algorithm.Then,the CC MFO algorithm is applied to the BP neural network,which is used for short-term prediction of network traffic.In the training of BP neural network,CCMFO algorithm is introduced to optimize.That the weight and threshold are adjusted by crisscross mechanism instead of the traditional gradient descent method,and the prediction accuracy of BP neural network is improved.The simulation experiment shows that the prediction model has good prediction accuracy and stability.Therefore,the improved moth flame algorithm proposed in this paper has certain practical application value.At the same time,the prediction results of this neural network are applied to the perception of the whole network sec urity situation,and a network security situation awareness method based on information entropy is proposed,which further improves the practicability of the neural network prediction model.Finally,the related research work is summarized.The next stage will further explore and improve the CC MFO algorithm and network traffic prediction model.That will be applied to more projects in the field of engineering to improve its application value.