Research On The Solution Of A 3D Packing Optimization Problem Considering Robot Packing Constraints

With the rapid development of my country’s economy and the progress of infrastructure and information technology,logistics is developing towards standardization,mechanization,automation and even intelligence.In the field of packaging,packing robots have broad application prospects.Packing is an important step in the packaging process of items,which belongs to NP-hard problem.With the increasing complexity of packing in real life,the constraints that need to be considered in the three-dimensional packing problem are also increasing.In order to use robots to improve packaging efficiency and reduce packaging costs,this paper proposes a threedimensional packing optimization problem considering the robot packing constraints with the goal of maximizing the loading rate.In addition,this paper also studies the model and solution algorithm of the three-dimensional packing problem considering the constraints of robot packing,and applies the obtained solution to the robot packing operation.The main work of this paper includes the following aspects:1.This paper analyzes in detail the three-dimensional packing optimization problem considering robot packing constraints.A mathematical model is established for the three-dimensional packing problem of multiple boxes,multiple items and multiple constraints under the condition of automatic packing by robots.The mathematical model takes the objective function of maximizing the loading rate.The mathematical model takes into account seven constraints.Among them,the robot packing constraints are mainly considered.The three-dimensional packing mathematical model of suckertype robot packing without considering robot packing constraints and the threedimensional packing mathematical model of splint-type robot packing considering robot packing constraints are established.2.This paper designs an ant colony algorithm based on the block generation intelligent algorithm,and uses this algorithm to solve the three-dimensional packing optimization problem.In order to solve the above model,this paper firstly customizes the ant colony algorithm according to the three-dimensional packing optimization problem from the aspects of initial parameter adjustment,pheromone update strategy and state transition rules.The ant colony algorithm based on the block generation intelligent algorithm is realized by adding the occupying point method,a certain coordinate adjustment strategy and the block generation intelligent algorithm to the ant colony algorithm.The algorithm can solve the packing layout.In addition,this paper uses the item loading strategy to sort the items to get the loading sequence.The packing layout and the loading sequence together constitute a solution to the three-dimensional packing optimization problem considering robot packing constraints.A threedimensional visualization of the solution to the three-dimensional packing optimization problem considering robot packing constraints is implemented in Python.3.This paper completes the action planning of the packing robot.In order to apply the solution of the three-dimensional packing optimization problem considering robot packing constraints to the robot packing,this paper uses the coordinate transformation method to convert the solution of the item-based coordinate system into the solution of the robot-based coordinate system.Then the path of the robot packing is modeled and solved.Moreover,the packing operations of the sucker-type packing robot and the splint-type robot are analyzed in detail.4.This paper solves the algorithm based on the case dataset and applies the solution to the robot packing.In this paper,the data set provided by a logistics company is used to solve the packing layout by the ant colony algorithm based on the block generation intelligent algorithm and the loading order.The average loading rate of the sucker-type packing robot that do not consider robot packing constraints reaches 0.833,which is better than that of related scholars;the splint-type robot that consider robot packing constraints still achieves an average loading rate of 0.772 when the splint consumes a certain space.The average loading rate is ideal.In addition,this paper takes the representative orders as an example to do the action planning of the sucker-type packing robot and the splint-type robot.