Prediction of Ammunition Storage Reliability Based on Improved Ant Colony Algorithm and BP Neural Network

Algorithm 1

IACO-BP (improved ant colony optimization Bp).

Initialize the ant colony parameters;

 begin:

  while(ants_i<ants_num)

   _i chooses paths based on pheromone concentration;

     ants_i ++

       //Select different strategies in stages

     if(iterations_num<N1)

       pheromones according to pheromone pure increase strategy

     if(N1<iterations_num<N2)

       pheromones according to pheromone accumulation and

volatilization strategy

     if(iterations_num>N2)

       pheromones according to pheromone doubling strategy

   if (current_error>margin_error)

        ants_i=1;

goto begin;

   else

      Input ant colony result is the initial solution of neural network;

      neural network training;

      resulte = neural network training results;

   return resulte;