Application of support vector machine optimized by improved ant colony optimization algorithm in power coal blending prediction

For current situation of power coal blending prediction, this paper proposed a new prediction model based on support vector machine(SVM) optimized by ant colony optimization algorithm(ACO). Firstly, use improved ant colony algorithm to optimize the parameter C of SVM and parameter sigma of kernel function; secondly, use the obtained parameters to build ACO-SVM prediction model; finally, applied the proposed model to the actual example. The verification results show that the proposed ACO-SVM prediction model can achieve higher prediction accuracy than basic SVM prediction model and Weighting average method under the condition of small samples; and it has a high practical value and broad application in coal blend field.