Application Of Genetic Algorithm In Making Blood Collection Plan For Hebei Blood Center

One of the most significant steps in the daily work of blood center is collecting blood.Drawing up blood collection plan is a process of arranging blood collection sites,teams,vehicles and personnel reasonably according to the needs of work.It is actually a problem of resource allocation under multi-constraints.However,the blood collection plan is formulated manually at present,consuming time and manpower.Furthermore,the established blood collection plan lacks rigidity,which hinders the motivation of employees.Because of many constraints,manual work can only make plans for the next three days.The uncertainty of the plan is not conducive to the rational planning of blood donation time for blood donors,thus affecting the daily blood collection and the guarantee of clinical blood use.Therefore,it is very important to use computer automation to formulate reasonable blood collection plan to reduce human time consumption.Diverse blood collection plan are actually combinations of different personnel,so the optimal conclusions are screened out in discrete sets.Traditional optimization algorithms have significant limitations.They can only select the optimal solution in the neighborhood of a certain point.They are powerless to solve the global optimal problem and require high differentiability of the objective function,which also limits the application field to a great extent.In this paper,genetic algorithm is selected to optimize the design through synthetical consideration of three algorithms including ant colony algorithm and annealing algorithm which have global searching ability.The main work includes:(1)Define the fitness function.According to the practical business model,the necessary and nonessential constraints affecting the quality of blood collection plan are determined,and then the fitness function which is suitable for the intelligent search of the optimal results by genetic algorithm is constructed.By investigating the actual working conditions,the weights of nonessential constraints are determined to make the objective function more accurate.(2)Gene coding.The blood sampling plan is coded by binary coding method which accords with the characteristics of chromosome genes.The gene of an individual in the population is composed of whether a doctor or nurse works at a certain place on a certain day,1 means to arrange work,0 means not to arrange work.According to the coding rules,the specific form of fitness function is further determined.(3)Algorithm implementation.In order to avoid prematurity,the first mutation operation is performed on the completely randomly generated initial population,and the first generation population of genetic algorithm is obtained.Then,the fitness function is used to calculate the fitness of each individual,and the Chromosomes for inheritance equal to the number of initial populations selected with corresponding probabilities.The crossover and mutation operations are carried out according to a certain probability in order to find the global optimal solution.The experimental results of the algorithm agree with the intuition.(4)Setting up automatic scheduling system of blood center.According to the algorithm,a blood center automatic scheduling system including database and underlying algorithm is further built,and intelligent scheduling operation is completed in this system.Through practical business application,the result is feasible.