Transfer robots (TRs) all have broad application prospects in industrial production or daily life, which can intelligently transport goods from one place to another, such as molten steel sampling robots, express delivery robots, and household service robots. The path planning and stable control of the above robots can be solved by using Genetic Algorithm (GA), because it is a Traveling Salesman Problem (TSP). However, the GA often suffers from the issues of slow convergence and low accuracy. Therefore, a path planning algorithm for TRs is presented, which applies the framework of GA and includes greedy population initialization, dynamic crossover-mutation probabilities, and Grey Wolf Optimizer (GWO). An intelligent controller is proposed to control TRs to move stably along the designated trajectory, which is combined by PID controller and feedforward neural network. Simulation results show that the presented path planning algorithm demonstrates strong optimization performance, which is better than GA. In addition, the provided intelligent controller with PID and feedforward neural network can make the controlled TRs move forward with high precision along with the planned path. Trajectory planning and high-precision control for arms of TRs will be our future work.