To ensure the safety and stability of automatic parking for intelligent vehicles in multi-vehicle scenarios, a cooperative parking system was proposed. The system enabled automatic parking of multiple vehicles while ensuring cooperative obstacle avoidance. The Hybrid A * algorithm was improved by incorporating vehicle steering constraints and inter-vehicle collision constraints during path search, enhancing motion safety and stability. The systems were categorised into parking system 1 and parking system 2 based on whether cubic spline interpolation was applied. The technique for order preference by similarity to an ideal solution (TOPSIS) method was used to select evaluation indicators and assign weights, quantifying the parking performance of different systems. Normalised scores were obtained to evaluate system performance. Simulation results under two-vehicle cooperative parking-in and parking-in/out scenarios indicate that parking system 2 consistently achieves higher scores than parking system 1. In the simultaneous parking-in scenario, one vehicle's score increases by 0.984 and the other by 0.420 2. In the parking-in/out scenario, both vehicles' scores increase by 0.420 2. The results demonstrate that parking system 2 performs better, effectively improving the safety and stability of multi-vehicle cooperative parking.