To solve the problem of multi-robot formation assembly point optimization under energy consumption and time constraints

a two-layer path optimization algorithm is proposed

which combines particle swarm optimization（PSO）and continuous Hopfield neural network（CHNN）.For the time constraint

a PSO algorithm is designed to solve the optimal formation set point.For the total energy consumption constraint

the Dubins path is used to transform the path planning into an assignment problem

and the optimal distribution path is searched based on CHNN.The numerical simulation results of three isomorphic underactuated robots show that the algorithm improves the search ability and can achieve the optimal formation aggregation.