经典的优化理论大多是在已知条件不变的基础上给出最优方案（即最优解），其最优性在条件发生变化时就会失去．局内问题与竞争算法则是针对特定的优化问题来研究这样的方法，它在变化因素的每一个特例中都能给出一个方案，使得这一方案所得到的解离最优方案给出的解总在一定的比例之内．本文应用复位策略给出限制图上局内ｋ 出租车调度问题竞争比为１＋ （ｎ－ ｋ）λ的竞争算法． The classical optimization theory mostly gives the optimal solution (that is, the optimal solution) on the basis of the known conditions, and its optimality will be lost when the conditions change. The intra-agency problem and competition algorithm are based on a specific optimization problem to study such a method, which gives a solution in every special case of the variation factor so that the solution given by this solution is the solution given by the optimal solution Always within a certain percentage. In this paper, we give the competition algorithm with the competition ratio of 1 + (n-k) λ for the k-taxi scheduling problem in the game.