Recent aging populations and labor shortages have accelerated the adoption of autonomous mobile robots and Personal Vehicles like electric wheelchairs in pedestrian spaces. However, the coexistence of diverse robots creates unprecedented safety risks including collision accidents. This necessitates architectural planning that ensures safe passage for all users in shared pedestrian-vehicle spaces. This study aims to classify avoidance behavior patterns based on crowd members' position coordinates and crossing feasibility between personal vehicles when two vehicles traveling in tandem approach a six-person crowd, as a foundational step toward human-robot coexistence. The research contributes to universal design through crowd flow simulation technologies, potentially reducing congestion and collision risks from the planning stage. Experiments were conducted in a laboratory where two personal vehicles intersected with a six-person crowd, recorded using fixed-point cameras. Experimental conditions included three variations each of intersection angle and inter-vehicle distance, with constant vehicle speed, allowing only crowd members to perform avoidance behaviors. Recorded video data were converted into sequential frame data at minute intervals to obtain time-series position coordinates in real space. Relative position and velocity vectors were calculated for each crowd member at their closest approach to the leading vehicle and visualized through graphs. Results showed that crowd members individually assess crossing feasibility between approaching personal vehicles. When inter-vehicle distance was short, strong tendencies emerged for avoidance in front of the leading vehicle. Conversely, increased inter-vehicle distance shifted the relative position vector distribution backward, with crowd members showing increased tendency to pass between vehicles. Short inter-vehicle distances provide insufficient space for safe passage, leading to preemptive avoidance maneuvers in front of the leading vehicle. Longer inter-vehicle distances create adequate passage space, enabling crowd members to judge crossing as feasible and successfully pass between vehicles.