路网结构的复杂性和运输生产的动态性 ,决定了要获得较大规模的实际空车调配问题的最优解是不可能或非常困难的。为求解此类问题 ,提出了一种有效的遗传启发算法。针对该问题的特定领域知识 ,设计了适合空车调配问题的矩阵编码方法 ,基于该编码的初始化方法能保证初始种群的可行性。针对矩阵编码提出的特殊交叉算子和变异算子 ,能保证生成的新个体总是有效的。通过对收发平衡和收发不平衡的空车调配问题的数值计算和分析 ,验证了所设计算法的有效性 The complexity of the road network structure and the dynamics of transport production determine that it is impossible or very difficult to obtain the optimal solution of the actual empty-vehicle deployment problem on a large scale. To solve such problems, an efficient genetic heuristic algorithm is proposed. Aiming at the domain-specific knowledge of this problem, a matrix coding method suitable for the empty vehicle deployment is designed. The initialization method based on this code can ensure the initial population feasibility. The special crossover operator and mutation operator proposed for matrix coding can guarantee that the new individual generated is always valid. Through the numerical calculation and analysis of the unbalanced receiving and sending and receiving and dispatching, the validity of the designed algorithm is verified