探讨了将遗传规划这一新的优化方法应用于巷道维护费用预测中。由于影响巷道维护费用的因素多，且各因素之间关系不明确，而遗传规划特别适用于各影响因素之间因果关系不明确的复杂非线性问题。因此，为预测巷道维护费用提供了一条新的技术途径，通过验证，证明该方法是可行的。 The application of this new optimization method of genetic programming to prediction of roadway maintenance cost is discussed. Due to many factors affecting the cost of roadway maintenance, and the relationship between the various factors is not clear, and genetic programming is particularly applicable to complex non-linear problems of causal relationship between the various factors is not clear. Therefore, it provides a new technical way to predict the maintenance cost of roadway, which proves that the method is feasible.