摘要 |
PURPOSE:To reduce the capacity of a storage device by computing the difference between the preceding value and the current value of a floating-point data sample value and compressing the mantissa part of difference data in the case of no change of the characteristic of difference data. CONSTITUTION:Process data of a process 1 is inputted to a process data input 2 and is converted to floating-point data by a data converting means 3 and is compressed by a data compressing means 4 and is stored in a process data base 6 of a storage device 5. If the difference between the preceding value and the current value of floating-point data is computed and found that the characteristic is not changed, the compression rate is changed in accordance with the value of difference data to compress the mantissa part of difference data, and difference data is stored. Thus, the floating-point data bit width is compressed to reduce the capacity of the storage device. |