二元函数的一个磨光算子

A Polishing Operator of Binary Function

给出了一个新的二元函数磨光公式，与文［１］中所给出的逼近阶０（ｈ２＋ｌ２）的公式相比较，逼近精度大大提高，在相同节点及计算量情形下，达到０（ｈ４＋ｌ４）阶，具有保凸性质。

A new polishing operator of binary function is presented. By comparison with the approximate degree 0(h2+12) of the formula presented in [1], the accuracy is raised so highly that in the condition of not inereasing the nodes and the amount of computation, the approximate degree can be achieved to 0(h2+12), and the convexity can be guaranteed.