JPH03231372A - Automatic wiring method - Google Patents

Abstract

PURPOSE:To execute the automatic wiring in which a path detection rate is high and a memory is used efficiently by preparing plural prescribed procedures, executing a search of high linearity and a flexible alternative search, and also, containing a path by a data structure of a binary tree node. CONSTITUTION:A procedure for executing a linear search from a source point and recording a coordinate of its point to a binary tree node at the time when an obstacle is detected, a procedure for searching an alternative route of the obstacle and detecting an alternative point, a procedure for executing a linear search and calculating a point coordinate at the time of reaching the next obstacle or a target line, a procedure for adding a binary tree node to this point and recording a coordinate of the point and a relation to the existing binary tree node, a procedure for back-tracing the data of the existing binary tree node at the time of reaching a new fault in an alternative point detecting process and changing a tree structure, a procedure for starting from an arrival point and executing recursively and repeatedly the second - fifth procedures until the time when the alternative search is ended, and a procedure for selecting an optimal path at the time when the alternative search is ended are provided and executed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an automatic wiring method, particularly in a CAD system for electrical systems, such as printed circuit boards, hybrid ICs, L
This invention relates to an automatic wiring method for determining wiring layouts such as SI.

[Conventional technology]

The automatic wiring method uses a computer to search for connection paths between component terminals and external connection terminals on a printed circuit board under various given conditions, and various methods have been proposed in the past. .

However, currently it is difficult to determine all the necessary wiring routes using automatic wiring, and generally the unrouted nets left by automatic wiring are manually searched for, and some of the existing wiring is removed and re-routed. Wiring is being done.

One of the most basic automatic wiring methods known to date is the maze router, which prioritizes searching for the shortest route. This method divides the wiring area into grids and adds various weights to the grids to determine the optimal route.

The algorithm creates waves from the starting point grid, determines the weights of the surrounding grids so that the wave spreads out, and exhaustively searches all paths to the ending point grid. . This method can always find the shortest route if there is a route, but the search takes a long time and requires a lot of computer memory.

On the other hand, many heuristic methods have been proposed in which only simple routes are tried, rather than brute-force trials like the maze router. A typical method is the line segment search method, in which X is searched from both ends to be connected.

Line segments are generated sequentially in the Y direction, and when a line segment from the start point and a line segment from the end point intersect, the route is determined by tracing in the opposite direction from that intersection, and the wiring in the X and Y directions is ,
Wired to the front and back of the printed circuit board via through holes.

[Problem to be solved by the invention]

The conventional maze router method described above can always find a wiring route if it exists, but it has the disadvantage of requiring a large amount of storage capacity and processing time. Furthermore, since the route is searched for by force, the wiring pattern tends to be unnatural, which often makes subsequent wiring correction work more difficult.

On the other hand, the line segment search method, which is a heuristic method, performs a search centered on simple line segments and requires only storage of wiring end point information, which requires less storage capacity and processing time. However, since route searches are performed based on simple line segments, there are many situations in which wiring routes cannot be found in actual automatic wiring situations, and the drawback is that it is only effective when the bins are arranged in a fairly orderly manner.

An object of the present invention is to provide an automatic wiring method that has a higher route discovery rate than conventional line segment search methods, uses memory more efficiently, and does not require a large memory capacity.

[Means to solve the problem]

The automatic wiring method of the present invention includes a first step in which a straight line is searched perpendicularly from a source point toward a target line passing through a target point, and when an obstacle is found, the coordinates of the point at which the obstacle is reached are calculated and recorded in a binary tree node. and a second step of setting a left or right direction flag on the obstacle, which searches for a route that detours around the obstacle to the left or right from the obstacle arrival point, detects the detour point, and indicates that the detour has been completed. , a third step of performing a straight line search vertically from the detour point toward the target line and calculating the coordinates of the target arrival point when the next obstacle or the target line is reached; A fourth step is to add a binary tree node and record the coordinates of the target arrival point and the relationship with existing binary tree nodes, and if a new obstacle is reached in the process of detecting the detour point, A fifth step of backtracing the data of the binary tree node to change the tree structure, and recursing the second to fifth steps starting from the target arrival point until the detour search for all obstacles is completed. A sixth step is to repeatedly execute the route, and a seventh step is to select the optimal route by backtracing the tree when the detour search for all obstacles has been completed.

〔Example〕

Next, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a flowchart in SPD format of one embodiment of the present invention.

The automatic wiring method shown in Figure 1 consists of a first step S in which a vertical line from the source point to the target line passing through the target point is searched, the first obstacle is detected, and a binary tree node is set.
1, a second step S2 of detouring around an obstacle and calculating a detour point, and a third step S3 of searching a perpendicular line from the detour point to the target line to detect the next obstacle.
a fourth step S4 of adding a new node to an existing binary tree node; a fifth step S5 of changing the tree structure using a backtrace when an obstacle is discovered during the detour;
The procedure includes a sixth procedure S6 that recursively repeats each procedure from the second procedure S2 to the fifth procedure S5, and a seventh procedure S7 that determines the optimal route by back tracing.

FIG. 2 is a layout diagram of an example of a printed circuit board for explaining the operation of this embodiment, and FIG. 3 is a data structure diagram of a binary tree node horizontally constructed according to this embodiment. The operation will be described in detail below according to FIG. 1 with reference to FIGS. 2 and 3.

In FIG. 2, a search is performed along a vertical line from source point 1 to target line 3 passing through target point 2.

When an obstacle 4 is detected on this vertical line, the coordinates of this obstacle arrival point 5 are x2. y is calculated and recorded in the fault arrival point coordinates 20 of the binary tree node 15 shown in FIG. 3 (Sl).

Here, a recursive call main routine S6 for detouring around the obstacle is called to enter the detouring search.

First, a left side detour search is performed for the obstacle 4, and the coordinates of the left side detour point 6 are calculated, and the values are temporarily stored. At the same time, a direction flag indicating that a left side detour has been implemented is added to the information about the obstacle 4 (S2 ). Next, this left detour point 6
A vertical line to the target line 3 is searched using the new source point, and when the next obstacle 7 is detected, the coordinates of the obstacle arrival point 8 are determined as Y8 is calculated (S3). Here, a new binary tree node 16 is added to the tree structure, and the coordinates of the obstacle arrival point 8 are X8. A backtrace pointer 23 that records Y8 in the binary tree node 16 and associates the new binary tree node 16 with the existing binary tree node 15
．． Pointer N 15+ for each left detour pointer 21
N16 is recorded (S4).

Next, the recursive main routine S6 is called again, and the left detour point 9 of the obstacle 7 is detected (S2
). Further, a search for a perpendicular line to the target line 3 is performed, and when the target line 3 is reached (S3), a binary tree node 17 is added and the target line arrival point 10
The coordinates Xlo.

Ylo and pointer N16 are recorded (S4). Furthermore, the pointer N17 is recorded in the binary tree node 16, the left side detour search ends, and the right side detour search begins.

In the right detour search, first, the right detour point 11 is detected from the obstacle arrival point 8, a right detour direction flag is set on the obstacle 7, and a vertical line to the target line 3 is searched.

When the target line 3 is reached, a binary tree node 18 is created and the coordinate x of the target line arrival point 12 is
t2. yl□ and each pointer N, 6. N, , and are recorded respectively.

Next, a right detour search for the obstacle 4 is executed from the obstacle arrival point 5, a right detour point 13 is calculated, a vertical search is performed, and when the obstacle 7 is detected, a binary tree node 19 is created, and the obstacle arrival point 14 is Coordinates of X14. Y14 and pointer N15 are recorded, and pointer N19 is recorded in the right detour pointer 22 of the binary tree node 15. However, since the left and right direction flags have already been set for obstacle 7, it is determined that no further search is necessary, and one binary tree node has a pointer N17. N1g is recorded, a pointer N19 is added to the binary tree node 17.18, and all detour searches are completed. By using the detour direction flag in this way, it is possible to avoid unnecessary multiple searches.

After completing the left and right detour search for all obstacles,
Select and determine the optimal route by backtrace (S7)
.

Although not shown in Figure 2, if a new obstacle is reached in the process of detouring around the obstacle and detecting the detour point, the position of the already created binary tree node will be changed to that obstacle. The tree structure is changed to avoid objects and return to a point where a detour point can be detected (S5).

The data structure of the binary tree includes a left detour pointer 21 for left detour, a right detour pointer 22 for right detour, and a backtrace pointer 23 for backtrace to form a tree. Furthermore, each coordinate value is stored in the obstacle arrival point coordinates 20. With this storage method, it is possible to represent two line segments with one binary tree node, so the memory capacity is about half of the usual memory capacity, and the memory usage efficiency is also improved.

〔Effect of the invention〕

As explained in detail above, the automatic wiring method of the present invention
Since it performs a highly linear search, it does not generate unnatural routes like a maze router, and it performs a route search while flexibly detouring according to the detection of obstacles, so it is difficult to use with conventional heuristic methods. A high route discovery rate that is difficult to find can be obtained.

In addition, since the region width for route search is set and the routes existing within it are covered by the binary tree node data structure, it is easier to determine the optimal route, and the binary tree node data structure is reduced to one binary tree node. Since it can represent two line segments, it has the effect of halving the storage capacity of the conventional method.

[Brief explanation of drawings]

FIG. 1 is a flowchart in SPD format of an embodiment of the present invention, FIG. 2 is a layout diagram to which the present invention is applied, and FIG.
The figure is an explanatory diagram showing an example of a binary tree data structure used in the present invention. 1...Source point, 2...Target point, 3...Target line, 4,7...
・Obstacle, 5,8°14...[F arrival point,
Left side detour point for 6, 9..., 10.12...
・Target line arrival point, 11.13...
... Right side detour point, 15 to 1 ... Binary tree node, 2o ... Obstacle arrival point coordinates, 2
1...Left detour pointer, 22...Right detour pointer, 23...Backtrace pointer.

Claims (1)

[Claims] A first step is to perform a straight line search perpendicularly from the source point toward the target line passing through the target point, and when an obstacle is found, calculate the coordinates of the obstacle arrival point and record it in a binary tree node. or a second step of searching for a route that detours around the obstacle to the right, detecting a detour point, and setting a left or right direction flag on the obstacle to indicate that the detour has been completed; a third step of calculating the coordinates of the target arrival point when the next obstacle or the target line is reached by performing a straight line search perpendicularly to the target line;
A fourth step of adding a new binary tree node to this target arrival point and recording the coordinates of the target arrival point and the relationship with the existing binary tree node, and a new obstacle is reached in the process of detecting the detour point. A fifth step is to backtrace the data of the binary tree node that has already been created and change the tree structure, and the second to fifth steps are performed starting from the target arrival point to remove all obstacles. The sixth step is to recursively repeat the detour search until the detour search is completed, and the seventh step is to select the optimal route by backtracing the tree when the detour search for all obstacles has been completed. Features automatic wiring method.