The existing airbag buffer system of airdrop equipment is poor in applicability, and it is difficult to meet the future demand of airdrop buffer of large weight equipment with low impact and anti-rollover. To solve these problems, this paper designs a combined airbag suitable for a new generation of airdrop equipment, adopts the optimal Latin hypercube design test method for sampling, and constructs a polynomial proxy model aiming at maximum overload and maximum rollover angle based on these samples. The multi-objective particle swarm optimization algorithm is applied to optimize the structural parameters of the airbag on the basis of the agent model, and the optimized cushioning system is subjected to finite element analysis: the maximal overload of the optimized airbag cushioning system is reduced by 24.7%, which improves the buffer performance;the maximal rollover angle is reduced by 42.6%, which greatly improves the lateral stability, and greatly reduces the risk of rollover. The research results show that the airbag cushioning system designed and optimised in this paper meets the future demand of large load and rollover resistance, and provides technical support for the airbag cushioning system of future airdrop equipment.