To address the challenges associated with long cycle times, low efficiency, high manufacturing cost, and significant energy consumption in the optimization design of transformer, a multi-strategy improved particle swarm optimization algorithm has been utilized. This algorithm is used to optimize the parameters of amorphous alloy dry-type transformer (designated as AMDT) in combination with an optimization system developed on the Visual Basic 6.0 software experimental platform. During the particle initialization stage, the Logistic-Tent chaotic map is applied to improve the initial diversity of the particles. Additionally, the dynamic learning factor and the nonlinear dynamic inertia weight coefficient are developed to improve the local optimization accuracy and enhance its global optimization ability. The optimization of the SCLBH19.400/10 amorphous dry-type serve as a case study, the particle swarm optimization, quantum particle swarm optimization, adaptive particle swarm optimization, chaotic particle swarm optimization, and multi-strategy improved particle swarm optimization algorithm are used to optimize the parameters. The experimental results show that compared with the traditional artificial design scheme, the traditional particle swarm optimization algorithm, and the other three improved particle swarm optimization algorithm optimization schemes, the multi-strategy improved particle swarm optimization algorithm significantly and improve computational efficiency. It achieves a reduction in total loss associated with amorphous dry-type transformer by 15.41% and decreases the main material cost by 14.81%. These results substantiate the effectiveness and superiority of the multi-strategy improved particle swarm optimization algorithm.