Research On Dual Robot Assembly Cell Multi-Product Scheduling Method And Simulation

Dual robot assembly cell with multiple assembly stations and multiple part libraries can meet the needs of small and medium batches and personalized customization production.The constraints in the multi-product scheduling problem of multi-robot assembly cell are complicated and strong,such as the robot collisions,which may lead to robot damage and waste work cycles.Furthermore,an effective method to solve this kind of scheduling problem that considering these factors is also an important concern.Therefore,to take the constraint of robot collision avoidance and multiple products,this paper studied a scheduling problem that is the dual robot assembly cell multi-product scheduling problem(DRAC-MPSP).The main research work of this paper is as follows:(1)To deal with the problem of robot collisions which may occur in multi-product assembly scheduling in DRAC,a scheduling model with robot motion constraint was established to minimizing the cycle time and tool switching times,and the constraints of multiple assembly stations,multiple part libraries and multiple end-tools were taken into consideration as well.(2)The single objective and multi-objective Hybrid Estimation of Distribution Algorithms(HEDA)are proposed to solve DRAC-MPSP.In single objective HEDA,the probabilistic model for product sequence and robot selection sequence was established respectively and the update mechanism of the probabilistic model was improved to make the algorithm have a better guidance.Furthermore,local search operation,genetic algorithm's mutation operation and elite solutions reservation operation were added to improve the local search ability of the algorithm and maintain the diversity.In multi-objective HEDA,the local search operation and global mutation operation used in single-objective optimization are retained,and the non-dominant solution of each iteration is recorded in the external Pareto-archive.The probability model in multi-objective HEDA is updated by using the new non-dominant solution at every iteration.Finally,the validity of the above model and method is verified by an example.The model and method proposed can be applied to the scheduling problem of multi-product assembly cell consist of more robots.(3)A simulation software for robot assembly cell scheduling is developed on VS platform based on Coin 3D graphics library.The proposed scheduling algorithm is integrated in this software,which can realize 3D scene construction of assembly cell,robot point teaching,collision detection,scheduling optimization simulation,generation of robot offline-program and generation simulation results functions.The validity and reliability of the proposed algorithm is verified by a simulation example.