the amount of sampling points over 57(N>57), the accuracy 0.1. Considered A/D converter and CPU and instruction, rounding error in a byte and double byte of fix-point and floating-point are accurately bound.

• N>57；有0.；1级准确度。 对舍入误差结合A/D转换器、CPU以及指令的实际使用情况；做了在8位、16位单、双精度以及定点、浮点情况下的准确界定。