<p>This paper considers the warehouse shop robot path planning problem through the proposition of a novel algorithm, the seeker optimization algorithm (SOA). In response to the problem of occasional target unreachable and efficiency improvement in path planning, which is caused by the inherent defects of SOA and the characteristics of the warehousing system, an enhancement strategy for SOA is introduced. Firstly, for the problem of low optimization efficiency due to the high randomness of the initial population in SOA, the concept of multiple swarm optimization is introduced, and a corresponding algorithm based on Positive Feedback Bootstrap-Monte Carlo estimation is proposed to generate a high-quality initial population for SOA so as to accelerate the convergence of the algorithm and to improve the optimization efficiency. Secondly, the path search method are refined by incorporating a two-way search and designing a direction-oriented node expansion method with node correction for reducing the number of unreachable target and saving the search time, and further to improve the SOA algorithm optimization efficiency. Additionally, aiming at the collision avoidance problem in multi-robot cooperative work scenarios, dynamic topology constraints are introduced and implemented, accompanied by the design of a collision avoidance strategy, which integrating with improved SOA is applied to address multi-robot conflict-free planning in the warehouse shop. Finally, the simulation is done to verify the validity of the proposed method.</p>