Research On Workshop Scheduling Problem Based On Swarm Intelligence Algorithm

With the in-depth implementation of the Made in China 2025 strategy,China is developing in the direction of a manufacturing power.High-quality manufacturing represents a higher level of real economic development and is the core driving force of economic growth.As an indispensable part of the manufacturing industry,workshop scheduling is the key technology for enterprises to achieve efficient production and high reliability,so it is of great theoretical and practical significance to study it.Swarm intelligence algorithm is an algorithm developed by simulating animal behavior,and has become a common method for solving optimization problems in recent years.In this paper,three different types of workshop scheduling problems are studied,and the swarm intelligence algorithm is improved and integrated,and the effectiveness of the algorithm is verified by solving these three types of scheduling problems.The specific research work of this paper is as follows:1.Aiming at the problem of slow convergence of genetic algorithms(GA)in the later stage,it is mixed with the Beetle Antennae search algorithm(BAS)with faster convergence,and a genetic longhorn beetle whisker mixing algorithm(BASGA)is proposed.Firstly,a multidirectional perceptual position feedback strategy is proposed for the poor optimization effect of the longhorn beetle whisker algorithm in high-dimensional problems and the insufficient exploration ability in the later stage,and the optimization ability and convergence speed of BAS under this strategy have been significantly improved.Subsequently,a hybrid strategy of GA and BAS is proposed,which uses BAS to process some individuals in the population to accelerate the convergence speed.Finally,the real number coding is adopted and converted to the process code by SPV rule,the appropriate genetic operator is selected,the solution method of the job shop scheduling problem is designed,and the test is carried out by example,which proves that the hybrid algorithm takes into account the optimization ability of GA and the convergence speed of BAS,and is effective in solving JSP problems.2.The parameter adaptation improvement of the particle swarm algorithm is carried out,and a particle swarm algorithm with multiple swarm mixing is proposed in combination with the algorithm in Chapter 3.The algorithm performs global search through multiple marginal populations performing BASGA,and the resulting solution is merged into the central population to participate in the operation of the improved particle swarm algorithm,and there is a migration phenomenon between each population to achieve the sharing of excellent individuals.According to the characteristics of FJSP,the chromosomes are divided into two segments,process segment and machine segment,and the Tent chaotic mapping is used to generate random numbers as the initial process segment,and the machine segment is initialized by global selection,local selection and random generation proportional mixing.New genetic operators were designed to enrich population diversity;The critical path and variable neighborhood search algorithms are introduced,and four neighborhood structures of FJSP are designed,and when the particle swarm algorithm runs to the later stage,the variable neighborhood search is performed to enhance its ability to jump out of the local optimal solution.Standard studies and actual cases are solved by algorithms to verify the improvement effect of algorithms.3.The problems related to dynamic flexible workshop scheduling were studied,including its concept,classification,research methods and related technologies.For the dynamic event of order insertion,the dynamic event is divided into several static windows through the dynamic event-driven strategy and rolling window technology,and the corresponding rescheduling scheme is generated by using the method of complete rescheduling and multiple swarm mixed particle swarm algorithms.Aiming at the dynamic event of machine failure,a partial rescheduling method is designed,and a combination scheme is formed with complete rescheduling,and when the machine fault occurs,the adaptive selection of the scheme is carried out according to the evaluation index.Finally,through the simulation experiment of FJSP,the feasibility of the algorithm in solving the dynamic scheduling problem is verified.