Optimal Allocation Of Counter Terrorism Forces Based On Genetic Algorithm

The study on the distribution of forces before, is usually considering the army’scombat capability as an average operational capability of the unit. But in fact,everyone’s capacity is strength difference. Especially China has developed peacefulfor decades, since the self-defense war against Vietnam, we have no large-scale warexperience, and great changes have taken place in the soldier quality, lack of profoundunderstanding of the actual combat. Therefore in the war of terror, especially in thecity, taking into account the actual capacity of each soldier, reasonable task allocationcan effectively improve the operation success rate and reduce the casualty rate. Thispaper considers the detachment (city) of CAPF counter-terrorism forces allocation.Genetic algorithm is a random search algorithm using the natural biologicalevolution. It can be used to guide the evolution according to the objective function ofthe individual and not to consider other information too much, Very practical forlarge-scale, nonlinear discontinuous and non-analytic expression of the functionoptimization. In practice, it can be widely applied to the fields of scheduling problem,image processing, automatic control, artificial intelligence and machine learning. Butthe standard genetic algorithm does shortcomings about global convergence,premature convergence and the random roaming etc., especially for multi-constraintsand multi-objective problem. The genetic algorithm has a great improvement since itproposed. In this paper, on the basis of previous work and combine with the actualproblem, we improved the processing of penalty function and constraints.The main work of this paper:1. Improved the penalty function style. In general, for the chromosomes that donot meet the constraints, we subtract the penalty function after the fitness function.But for a class of constraints, expression of each gene is associated with it; we canimprove the penalty function with the gene and to calculate the fitness later. This can have a greater probability of retention of good gene in the crossover and mutationoperation.2. Combine the constraints with the crossover and mutation operator. For theconstraints that associated with the chromosome entirety, we can design the operatoraccording to the constraints to support the finally solutions which are all feasible.3. Design and realization the soldiers’ strength database and force allocationsystem. In this system, given the number of dispatch and the rendezvous, it can workout the optimal solution according to the database and it can provide reference fordecision-making for commands.