Embedded Control System Design Of Food Delivery Robot

With the continuous advancement of science and technology,robots are playing an important role in people’s lives,and the catering industry has also begun to use robots to deliver food,but at present,food delivery robots are often used in indoor scenes such as restaurants and hotels,and the outdoor delivery of these campuses still requires manual delivery by delivery staff,which is laborious and laborious,and there are problems such as the wrong location and wrong take-out delivery.To solve the above problems,an outdoor food delivery robot scheme is proposed,when a customer orders,the merchant places the food in the meal locker,and the food delivery robot control system plans the best delivery path and delivers the food safely to the target location.According to the application requirements and functional goals of the food delivery robot,the embedded control system of the food delivery robot is designed,and the autonomous navigation is regarded as the key link in food delivery,and the path planning based on laser SLAM and the improved A* algorithm is studied in depth.First,Solidworks was used to build a structural model of the food delivery robot,and in order to facilitate the safe distribution of food in the campus environment,a six-wheel wheel structure was adopted.Then,the underlying control system based on the STM32 main control chip is built,the hardware circuit of each module of the control system is designed,the functional modules such as motor drive,obstacle avoidance,environmental perception and communication are designed,the motor and sensor used are physically selected,and finally the modules are summarized to build a complete circuit design.On the basis of hardware,drivers are written for each module to complete software development for motor drive,obstacle avoidance control and communication.Second,focusing on the path planning based on laser SLAM and improved A* algorithm,the model built by Solidworks is first simplified,and the URDF model file of the food delivery robot is written to pave the way for subsequent visual testing.Food delivery robot laser SLAM uses gmapping algorithm to build an environment grid map,the real-time positioning of the robot in the map selects the AMCL positioning method,and the move＿base package used in path planning is developed and configured,the function package global path planning adopts A*algorithm,the planning takes a long time,the path turning point is more,not smooth enough.Aiming at the problems existing in path planning,the following optimization improvements are proposed for the A* algorithm: the Manhattan distance is used as the heuristic function of the A*algorithm to reduce the number of nodes explored by the planning;The weight ratio coefficients of the actual cost g(n)and the estimated cost h(n)in the evaluation function f(n)were adjusted to improve the efficiency of global path search.Add a second-order Bezier curve to curve the inflection and inflection point locations in the planned path to make the path smoother.By comparing multiple sets of experiments in matlab,it can be found that the improved algorithm path planning takes less time and the number of nodes explored is small,which improves the search efficiency.Then,Gazebo is used to build an outdoor simulation environment,the URDF model of the food delivery robot is introduced in Gazebo and RViz,the raster map required for food delivery is established by laser SLAM,and the map is saved to RViz to simulate and verify the positioning and path planning of the food delivery robot,the results show that the food delivery robot can well realize path planning and autonomous navigation,and compare the A* algorithm with the improved A* The results of the path planning of the algorithm in the virtual environment can verify the superiority of the improved algorithm.Finally,the underlying control of the food delivery robot is tested,and the lidar is used to map and navigate the actual environment.