Complex Branch Pipe Layout Optimization Based on Machine Learning and Improved Grey Wolf Algorithm

Abstract

Abstract Branch pipes are widely found in complex equipment, and its layout planning and design has an important impact on the quality and efficiency of equipment design. In this paper, a branch pipe layout planning method based on machine learning and improved grey wolf algorithm is proposed. Firstly, the branch pipe layout problem is decomposed into a series of “between two points” pipe layout and “point to line” pipe layout problems to reduce the complexity of the problem. Secondly, the Q-learning algorithm is applied to solve the “between two points” pipe layout path solving problem, and then combined with the improved grey wolf algorithm to optimize the “point to line” pipe path solving. Finally, a simulation system is built based on Siemens NX and MATLAB platform, and the validity of the proposed method is verified by simulation cases.