In order to improve the accuracy and intelligence of the mowing robot autonomous navigation and localization, design a model of mowing robot based on FPGA ( field programmable gate array) and the neural network algorithm, the design using FPGA reconfigurable technology, to three layer back-propagation neural network as a typical model. Using BP algorithm formula, the design of the Intelligent Robot Mower control model. By using the FPGA technology, design the hardware system of the mowing robot. Finally, the design method of the input text, using field experiment, the robot trajectory planning ability and control precision was verified. The experimental results show that using FPGA and the neu-ral network model can effectively through five obstacles, and can get the satisfactory results of trajectory planning. Com-pared to the ordinary PID controller and neural network PID controller control the error, results can be seen from the, neural network PID controller of the mowing robot control error to be significantly better than the traditional PID controller error small, the method for neural network hardware implementation provides a reliable theoretical basis.