Research And Application Of Intelligent Rectangular Packing Problem Based On Cuckoo Search

At the historical juncture of a new round of scientific and technological revolution and industrial transformation,reducing the consumption of materials in the nesting link is of great significance for improving the economic efficiency of enterprises and reducing emissions and consumption in society.Firstly,taking the rectangular packing problem as the research object,the background,purpose and significance of this paper are expounded from the perspective of the development of heavy machinery industry and green production.Then,from the perspective of algorithm,the paper summarizes the research progress of rectangle layout at home and abroad,and analyzes the development status and research trend of transfer learning technology.Then,the basic rectangular layout model is constructed,its process characteristics and calculation difficulties are analyzed,the principle and basic flow of the excellent algorithm are introduced in detail from the two aspects of sequencing and positioning algorithm,and its shortcomings and future research direction are summarized.Finally,the research content and framework of this paper are given.Secondly,the mathematical model of the rectangular packing layout problem corresponding to the nesting of medium thickness steel plate is abstracted,and the basic theory of cuckoo search optimization algorithm is introduced,and the technical difficulties in solving the discrete combination optimization problem are analyzed.Based on this,this paper redefines and discretization the intelligent behavior of cuckoo algorithm,designs the decimal coding method and the rules for calculating the nest distance,and proposes the levy flight mechanism based on the learning and neighborhood search operator and the nest parasitic update mechanism based on the disturbance factor.In order to overcome the shortcomings of the algorithm of lowest horizontal line,the algorithm of lowest horizontal line based on value evaluation is proposed.The effectiveness and practicability of the new discrete cuckoo search algorithm are verified by the international standard test set.Moreover,abstract the coil nesting problems with rectangular pieces of a corresponding mathematical model,in order to achieve the reuse of knowledge,improve the quality and efficiency of the sample solution,this paper introduces the Q-learning and knowledge transfer to discrete cuckoo search algorithm,put forward Transfer Cuckoo Optimizer algorithm.In this algorithm,the high-dimensional space reduction method based on chain extension is used to avoid the dimension explosion of knowledge space.In order to explore and utilize the knowledge of equilibrium population,the labor division and cooperation model of cuckoo population was introduced on the basis of probabilistic search,and a new strategy of action selection was proposed.After the completion of the pre-learning stage,the optimal knowledge matrix is transferred to the target task through the dual-source linear migration strategy,which helps to guide it to obtain the layout result quickly.The experimental results show that the TCO algorithm designed in this paper not only improves the quality of the solution of the layout problem,but also can achieve 1.5～4 times of the solution speed of other algorithms.Finally,the functional modules of the steel plate nesting system are designed based on the actual demand,combining with the working process of the steel plate nesting process and the optimization of the layout requirements.Then,by using the commonly used system development technology and combining the above contents,a steel plate nesting optimization system is developed.Finally,an engineering example is used to verify the practicability and effectiveness of the steel plate nesting system.