A GABP optimized algorithm for filler rate of non-heated wire

For the process of submerged arc welding filled with non-heated wire, the filler quantity of non-heated wire is one of the main parameters that affects the microstructure and mechanical properties of welded joints.First, through a lot of welding experiment, the effect of the filler quantity of non-heated wire on the microstructure and mechanical properties was investigated.On the basis of the heat balance law in welding process, the balance equation of heat dynamic distribution for the process of submerged arc welding filled with non-heated wire was established.Then the relational expression was formulated for the filler rate of non-heated wire.At last, the nonlinear mapping relationship between the welding current I, arc voltage U, welding speed v and filler rate of non-heated wire vl was realized based on artificial neural network.Experimental results showed that, the optimized algorithm of BP neural network based on genetic algorithm(GABP) can realize adaptive control for the process of submerged arc welding filled with non-heated wire.The linear correlation between filler rate of non-heated wire in actual welding process and the expectative output comes up to 0.991 88.This shows that the algorithm of GABP can meet requirement for welding process and properties.