SOURCES OF ERRORS WHEN CALCULATING MATHEMATICAL MODELS OF COMBAT ACTIONS IN COMPUTER SIMULATION SYSTEMS

Currently, there are many situations when practically conducting a scientific experiment or recreating this or that situation in order to make an informed decision is impossible or requires significant financial costs. In such cases, computer simulation systems always come to the rescue, which have been used for a long time to make decisions affecting human lives, planning military operations, directions for the development of the types of the Armed Forces and industries, and even the prospects of state formation. But specialists in the relevant fields, based on the results of computer calculations with the representation of numbers in binary floating-point code, do not take into account the peculiarities of the accuracy of calculations in a computer with binary numbers and the influence of binary arithmetic on the results of simulation modeling. As a rule, the mathematical results of computer calculations are considered to be reliable, although they may contain an error that depends on the hardware of the electronic computing machine, namely on the bit rate of the processor bus. This is due to the IEEE 754 technical standard for floating-point arithmetic. In the article, the authors analyzed the mathematical models used in combat simulation systems and approaches to their solution. The representation of binary floating-point numbers in computer systems of different bit sizes has been detailed, and possible sources of errors when numbers are represented in binary code have been determined, as well as mathematical calculations. This approach allows you to take into account possible error limits at the stage of entering data into computer systems for calculation, as well as directly when performing mathematical calculations in electronic computing systems.