Abstract: Robot automatic welding requires controlling multiple parameters, and the entanglement and combination of multiple variables explode and exhibit non-linearity. The complex search space makes it difficult to optimize multiple welding parameters, resulting in insufficient welding flatness. Therefore, a hybrid ID3 algorithm optimization method for robot automation welding process parameters is proposed. Construct a data sample set by collecting welding data on process parameters such as wire diameter and arc voltage. Introducing fuzzy theory and hesitant fuzzy set theory to improve the ID3 algorithm, capturing nonlinearity and fuzziness in multivariate data by considering multiple membership functions. Using the hybrid ID3 algorithm to construct a decision tree model, calculate the information entropy and information gain of multivariate quality index data, select the process parameters that have the greatest impact on the classification results of welding quality as splitting attributes, narrow down the search space, and reduce classification uncertainty. Analyze the impact of different parameter combinations on welding quality through hesitant fuzzy decision tree, and select the optimal process parameter combination. Experimental results have shown that this method can effectively construct a hesitant fuzzy decision tree for three welding quality indicators: weld width, weld depth, and residual height, and optimize the welding process parameters for robot automation; After applying this method, the welding flatness can be effectively improved.