一维平均散布随机数序列的发生方法(国外英文资料).docVIP

一维均匀分布随机数序列的产生方法(国外英文资料) From the above data, it is easy to see the defects of its easily reduced to 0. Times the same method: The iteration of the same method is as follows: ??????????? Each parameter meaning and the action of each step can be seen in 1 Of course, there is a theoretical basis for the selection of parameters here, otherwise the period of random Numbers generated will be smaller and more relevant. The coefficients of the iterative formula of the two groups that have been tested by the two groups are: 1); 2). The blending of the same method: The hybrid and residual method is a mixture of the congruence and multiplication, and the iteration is as follows: According to previous research, under the conditions under conditions, the parameters are guaranteed by the following selection, and the probability statistics are good: ? However, a careful study as you can see its a fatal weakness, namely the generation of random Numbers in the given week period into linear growth trend, obviously, in most cases, this kind of extremely rich regular type of random number should not be used. Inverse function: When the random variable is continuous, you can use the inverse method. ? By using the principle of probability integral, the distribution function F (X) of random variable X can be applied to the inverse function. In this case, a random variable of uniform distribution within a 0-1 interval. When F (X) is simpler, its easier to solve, and when F (X) is more complex, you have to use a more complex transformation technique. For the uniform distribution, heres an example: It is known that the projectile has a uniform distribution of the azimuth of the target, and the experiment (0, 1) uniform random number transformation is used to simulate the sampling value of the position Angle of the projectile. Solution: In this case, R is the random number of the uniform distribution on the 0-1 interval. Because this is a linear transformation of the distribution of 0 minus