A Method of Accelerating Kriging Interpolation
A technology of kriging interpolation and end-kriging, which is applied in the field of geology to reduce the search range, reduce redundant calculations, and improve the search speed.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
specific Embodiment approach 1
[0032] The specific embodiment one, a kind of method for accelerating Kriging interpolation described in the present embodiment is carried out according to the following steps:
[0033] Step 1. On the CPU side, sort the observation points from small to large according to the abscissa values of the known observation points;
[0034] Step 2, transmit the sorted data to the GPU, and start the Kriging interpolation program on the GPU side;
[0035] Step 3: In the GPU, each thread executes the calculation of a point to be estimated, and all threads first search for n known observation points x around the unknown point x to be estimated j ,j=1,2, … ,n, as the neighbor point of the unknown point;
[0036] Step 4. Synchronize all the threads in the same workgroup in the GPU, and execute the next step after the search of the neighboring points of all unknown points to be estimated in the workgroup is completed;
[0037] Step 5. The first thread in all workgroups compares the neigh...
specific Embodiment approach 2
[0049] Specific embodiment 2. This embodiment is a further description of a method for accelerating kriging interpolation described in specific embodiment 1. The search process described in step 3 is as follows:
[0050] 1. Compare the abscissa value of the point to be estimated with the abscissa of the observation points that have been sorted, and find its order in the observation points, that is, x i i+1 ;
[0051] 2. Define an array with a size of n to store n adjacent points of x, and start searching from the position of point x to the left and right sides alternately. There are two cases to deal with: 1. The array is not filled: the searched Points are filled into the array; 2. The array is already filled: if the searched point satisfies the following condition 1, replace the adjacent point in the array with the maximum distance from x, otherwise continue searching; if condition 2 or condition 3 is met, the search is completed;
[0052] Among them, condition 1 is: the di...
specific Embodiment approach 3
[0055] Specific Embodiment 3. This embodiment is a further description of a method of accelerated Kriging interpolation described in Embodiment 1 or 2. In step 5, the method of identifying unknown points to be estimated with the same adjacent observation points: as figure 1 As shown, each grid in the figure represents a point to be estimated, and A, B... respectively represent the set of neighboring points of the point to be estimated. If the adjacent points to be estimated have the same neighboring points, then the grid in the figure The letters in are also the same;
[0056] a. Compare each point to be estimated with the point to be estimated on its left side respectively (except the leftmost point to be estimated), if the same, it is represented as ← in the figure;
[0057] b. Compare each point to be estimated with the point to be estimated on its upper side respectively (except the topmost point to be estimated), if the same, it is represented as ↑ in the figure;
[0058...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More