<p>As the task environments faced by robots become increasingly complex, dense obstacles and potentially unknown obstacles in the environment pose a significant challenge to the robot’s path planning. In this case, unknown obstacles tend to be more menacing to the robot than known obstacles. In this paper, considering known and potential threats in complex environments, a hybrid path planning algorithm named PJA*-WPAPF is proposed, which can handle dense obstacle environments and deal with unknown obstacles in real time. To reduce the redundancy and complexity of paths generated by the traditional A* algorithm, a passability judgment strategy-based 8-neighborhood search is introduced to improve both search efficiency and path safety. Additionally, a selective insertion of control points strategy based on the B-spline technique is proposed to improve the smoothness and security of the global path. To address the local optimal solution of the artificial potential field algorithm, the weighted-distributed attractive points method is proposed, along with the obstacle evaluation strategy. Subsequently, the two improved algorithms are combined using an event-triggered mechanism, resulting in the PJA*-WPAPF algorithm. The algorithm was verified on a grid map with both known and unknown obstacles being incrementally introduced. Under such map conditions, comparisons were made between the PJA*-WPAPF algorithm and three other algorithms. The simulation results show that while the success rates and path quality of the other algorithms significantly decreased with the increasing environmental complexity, the PJA*-WPAPF algorithm maintained a success rate close to 100%, while ensuring excellent smoothness, safety, and cost-effectiveness.</p>