Research On Robotic Cells Hybrid Flow Production Scheduling Optimization Method

Robotic cells is a kind of advanced production system,it is widely used in textile industry,iron and steel industry,semiconductor industry and automobile manufacturing industry and so on.Since robotic cells is applied in many fields,robotic cells scheduling problem can be easily found.In addition,because of demanding from large quantity and small variety to small batch and multi variety,and requiring individuation and customization by customers,hybrid flow production is considered.In this thesis,robotic cells and hybrid flow production are combined,named robotic cells hybrid flow production,robotic cells hybrid flow production scheduling problem is proposed.As soon as robotic cells hybrid flow production scheduling problem is optimized,the utilization of existing resources is improved,the management level is enhanced,and competitive strength is increased,what is more,market demand which properties are fast variety is met,production quality,production price and production delivery time that are required by customers are satisfacted.In order to realize the above objectives,the specific research contents are as follows:The methods of optimizing robotic cells hybrid flow production scheduling problem is researched in the third chapter.Because the problem is NP-hard,computation time is much long if exact methods are used,especially big scale;the quality of solution is poor by heuristics methods,so chemical reaction optimization(CRO)is adopted.Improved chemical reaction optimization(ICRO)is developed.In ICRO,to obtain initial population,sequential insertion(SI)method is proposed.For realizing basic reaction operators,Random exchange,single point sequence crossover,random insertion and distance preserving crossover operator are applied.The results are better than the published results.In order to avoid premature of ICRO,chemical reaction optimization based on local search(CROLS)is provided.In CROLS,minimization of blocking time close part swap(MBTCP)is constructed for local search.Orthogonal experiment is designed for the parameters of CROLS.CROLS is better than ICRO by solving randomly generated stances.The performances of CROLS are better than the performances of ICRO.As soon as CROLS and ICRO are used,production efficiency is promoted,and output is raised.The methodology of solving robotic cells hybrid flow production scheduling problem with more tanks(more than three tanks)is considered in the fourth chapter.Robotic cells hybrid flow production scheduling problem is optimized according to following steps,firstly,the problem with three tanks is transferred to the problem with two tanks,secondly,optimum solution of the problem two tanks is researched,finally,the third robotic move is inserted the optimum solution.If the method which is used to solve robotic cells hybrid flow production scheduling problem is adopted in the fourth chapter,solving process is tedious,furthermore,it's hard to get a satisfactory solution.In order to resolve the problem,firstly,robotic activity which is shown for transferring double sequence into single sequence is defined.It is means that robotic move sequence and part input sequence are transferred robotic activity scheduling,and the difficulty of considered problem is reduced.Then properties of feasible solutions which evolutionary operator is designed by are investigated.Thirdly,feasible robotic activity insertion method(FRAIM)is addressed,feasible solution is constructed and double layers filtered variable width beam search(DLFVWBS)is realized by FRAIM.Finally,an effective chemical reaction optimization(ECRO)is proposed.In ECRO,a new encoding method,robotic activity encoding,is shown for transferring double sequence into single sequence.To obtain initial population more effectively,insertion robotic activity method(IRAM)is firstly addressed.Furthermore,for enhancing the efficiency of ECRO,elementary reaction operators are designed according to properties of feasible solutions.For escaping from local optimization solution,improved tournament selection is proposed.From the simulation results,compare to stochastic generated solution method(SGSM),IRAM is outstanding.The performance of ECRO is better than branch and bound(BB)method and DLFVWBS.After ECRO is applied,productivity per unit is improved,production costs is reduced,and profit is increased.The method of robotic cells hybrid flow production scheduling problem setup times and two tanks is studied in the fifth chapter.If setup times is considered,production cycle is lengthened,energy demand is increased,and production efficiency is reduced.It is valuable to research robotic cells hybrid flow production scheduling problem setup times.these published algorithms are easy to obtain locally optimal solution,A new variable neighborhood search(NVNS)is designed in the field.In NVNS,initial solution is generated by improved MM;two neighbourhood structures,random exchange and random insertion,are constructed;Parameters of NVNS is analysed for finding better solution.NVNS is compared to these published algorithms by solving randomly generated instances,the results show that the proposed algorithm is more effectively,and production cycle is optimized,production costs per unit is reduced,and management level is improved.