Variable ordering heuristics play a central role in solving constraint satisfaction problems. Combining two variable ordering heuristics may generate a more efficient heuristic, such as dom/deg. In this paper, we propose a novel method for combining two variable ordering heuristics, namely Pearson-Correlation-Coefficient-based Combination (PCCC). While the existing combination strategies always combine participant heuristics, PCCC checks whether the participant heuristics are suitable for combination before combining them in the context of search. If they should be combined, it combines the heuristic scores to select a variable to branch on, otherwise, it randomly selects one of the participant heuristics to make the decision. The experiments on various benchmark problems show that PCCC can be used to combine different pairs of heuristics, and it is more robust than the participant heuristics and some classical combining strategies.

中文翻译：

---

关于结合约束约束问题的变序启发法

变量排序启发法在解决约束满足问题中起着核心作用。结合两个可变排序启发法可以生成更有效的启发法，例如dom / deg。在本文中，我们提出了一种新的方法来结合两个变量有序启发式算法，即基于Pearson-Correlation-Coefficient-based的组合（PCCC）。现有的合并策略总是将参与者试探法组合在一起，而PCCC会在搜索上下文中将参与者试探法进行组合之前，先检查参与者试探法是否适合合并。如果应该将它们组合，则将启发式分数组合起来以选择要分支的变量，否则，它会随机选择参与者启发式之一来做出决定。针对各种基准问题的实验表明，可以使用PCCC组合不同的启发式对，