Automatic routing method and program

Abstract

The goal is to minimize manual correction work and achieve highly accurate results by eliminating the need for users to manually change automatic routing settings or perform programming tasks. [Solution] The system consists of the following steps: 1. Calculate candidate routes; 2. Evaluate whether the candidate routes conform to various design rules; 3. Calculate an overall evaluation index Y; 4. Repeat steps 1-3 to save the coordinate data of the candidate routes, the evaluation index Y, etc., to a performance database; 5. If, among the database-saved routes saved in the performance database, there is a route that is more similar to a user-modified route modified by a user than to the automatically routed route with the best evaluation index Y, modify the evaluation function, etc., so that the automatically routed route approaches the user-modified route; and 6. Save each data to a modification history database that can be referenced during the next routing.

Description

[Technical Field]

[0001] This invention relates to an automatic routing method and a program for executing it, which automatically creates design drawings and construction drawings for piping, ducts, wiring, etc., that connect equipment and other devices in building facilities and plant facilities. [Background technology]

[0002] Traditionally, large-scale buildings such as power generation facilities, various plant facilities, office buildings, and factories have required the placement of vast amounts of equipment, piping, ducts, cable racks, etc., within the building. When creating the design and construction drawings for these structures, a wide variety of rules are applied from the perspectives of reducing construction costs, ease of inspection and maintenance, and ease of construction. These design and construction rules include common sense rules such as connecting with the shortest possible routes and minimizing bends (joints), as well as rules specified in construction manuals to ensure various functions, and rules held by the client, building owner, or operator (hereinafter, these rules will be collectively referred to as "design rules"). Traditionally, skilled design engineers, construction managers, construction drawing company employees, CAD operators, etc. (hereinafter, these will be collectively referred to as "users") have had to consider each rule while creating the designs and drawings, which required a great deal of effort.

[0003] In recent years, CAD software for plant and building facilities (hereinafter referred to as "CAD software") has been developed that automatically routes piping, wiring, and ducts between equipment while satisfying design rules defined in the software, once each piece of equipment is laid out on a drawing. Applying these design rules to such CAD software requires algorithmization of those rules and programming work.

[0004] For example, Patent Document 1 discloses a route creation device comprising: a cell generation device that generates cells; a route search device that generates routes using a route search algorithm; a route selection device that selects routes using a route selection algorithm; a real-coordinate transformation device that converts routes into real coordinates; a dataset containing route information; a learning means for learning the dataset; and a learning operation means for utilizing the results learned by the learning means, wherein the results learned by the learning means are used for at least one of the route generation by the route search device and the route selection by the route selection device. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Patent No. 7462473 [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] In the aforementioned Patent Document 1, Must rules, which must be followed, and Better rules, which are desirable to follow, are defined. In the first stage, route search, numerous routes that comply with the Must rules are generated, and then in the second stage, route selection, the optimal route that complies with the Better rules is selected from among them. Furthermore, in the aforementioned Patent Document 1, in addition to route generation using a genetic algorithm, a deep learning-based learning method is used in conjunction to select the optimal route.

[0007] However, when design rules are numerous, they can sometimes conflict with each other, or the priority of rules may differ depending on the site and application. As a result, users may need to make modifications to drawings generated by automated routing.

[0008] On the other hand, improving the accuracy of the automatic routing results output by CAD software requires changes to the input conditions and program configuration of the CAD software. However, even if a user manually modifies a drawing, these changes are not reflected in subsequent automatic routing results (for example, automatic routing results obtained by redrawing a previously automatically routed drawing with changed conditions, or new automatic routing results created for a different site or project). As a result, drawings that require modification each time are output, and consequently, the benefit of reducing the time required for drawing is not achieved.

[0009] Therefore, the main objective of the present invention is to provide an automatic routing method and program that minimizes manual modification work and achieves highly accurate results without requiring users to manually change automatic routing settings or perform programming tasks. [Means for solving the problem]

[0010] To solve the aforementioned problems, the present invention according to claim 1 includes a first step of calculating candidate routes using a route search algorithm, The system checks whether the candidate route conforms to various mandatory rules that must be followed, and also checks to what extent the candidate route conforms to various evaluation rules that determine superiority or inferiority based on predetermined evaluation perspectives, using conditions predetermined by the engineer or user. or repetitive calculation formulas, arithmetic expression The evaluation element function F is composed of the above and takes coordinate data as input. i Evaluation element index x calculated from i The second step involves scoring using, The aforementioned evaluation factor index x i and the influence coefficient a that weights it i A third step involves calculating an overall evaluation index Y of the candidate routes using an evaluation function G that has the above as its components and determines the overall superiority or inferiority taking into account each of the aforementioned evaluation rules, Repeat the above first step to third step, and for each iterative calculation, the coordinate data of the candidate route calculated, and the influence coefficient a i associated with it, the evaluation element index x i and the evaluation index Y are stored in a candidate route database in a fourth step; As a result of performing iterative calculations in the above fourth step, based on the evaluation index Y, the best candidate route is used as the automatic routing adopted route, A route modified by the user as needed for the automatic routing adopted route is used as the user modified route, When the route stored in the candidate route database is used as the candidate database route, Search whether there is a route in the candidate database route that is more similar to the user modified route than the automatic routing adopted route. If there is a similar route, the influence coefficient a i is adjusted so that the automatic routing adopted route approaches the user modified route, i and a fifth step of modifying at least one of the evaluation element function F and the evaluation function G is added; i The influence coefficient a i modified or retained in the above fifth step, the evaluation element function F i and the evaluation function G, the user information of the operator who performed the operation, the coordinate data of the user modified route, and the coordinate data of the route in the candidate routes that is most similar to the user modified route, the influence coefficient a i and the evaluation element function F

[0011] are stored in a modification history database to enable reference to each of the above data during subsequent routings, and an automatic routing method characterized by including a sixth step is provided.In the invention described in claim 1 above, in the first step, candidate routes are calculated by trial and error based on probability theory, such as conventionally known genetic algorithms. In the second step, the candidate routes are checked and scored to determine whether they conform to various design rules predetermined by an engineer (mainly an automated routing software developer) or a user (the "user" mentioned above, mainly a person who creates design drawings and construction drawings). Next, in the third step, an overall evaluation index Y of the candidate routes is calculated. Steps 1 to 3 above are repeated for various candidate routes, and the results are stored in a database for candidate routes (fourth step).

[0012] The reason for the difference between the automatically routed route obtained in the above flow and the user-modified route modified by the user is that the priority of various design rules differs between the automatically routed route and the user-modified route, or the evaluation factor index x i Evaluation element function F i This is thought to be due to inappropriate settings or calculation formulas for the evaluation function G. Therefore, in the fifth step of the present invention, the modifications made by the user to the automatically routed route are used as training data (correct answers), and a function is added to feed back the user's modification intentions to each evaluation element of the automatically routed route, with the goal of reducing the difference between the modified route (correct answer) modified by the user and the route that the user is likely to modify, using the modified route (correct answer) modified by the user as the criterion for judgment.

[0013] Specifically, the system searches the candidate route database for routes similar to the user-modified route rather than the automatically routed route, and if a similar route is found, it adjusts the influence coefficient a so that the automatically routed route approaches the user-modified route. i , the evaluation element function F i And perform an operation to modify at least one of the evaluation functions G. As a result, in the second and third steps, the evaluation element index x iThe scoring and overall evaluation index Y will reflect the user's intentions, and highly accurate automated routing results will be obtained without the user having to manually change settings or modify programs each time automated routing is performed.

[0014] Furthermore, by providing the sixth step, which saves and stores the calculation results up to the fifth step above, as well as the user's modifications to the drawing and the user's information, in a modification history database, and allows these values ​​to be referenced in the third step, the convenience of the automatic routing function for users is improved, and new users can also share and utilize it.

[0015] As part of the present invention according to claim 2, the automatic routing method according to claim 1 is provided, which includes one or more of the following essential rules: the target route is connected without interruption between the designated start and end points; it does not interfere with other routes or obstacles; the route between the start and end points is connected as horizontally and vertically as possible; and in routes where a horizontal gradient is required for horizontal pipes, a gradient within a predetermined range is ensured.

[0016] The invention described in claim 2 above lists specific items of mandatory rules that the candidate route must follow.

[0017] As part of the present invention according to claim 3, the automatic routing method described in claim 1 is provided, in which the evaluation rule includes one or more of the following: the length of the route is short, the number of bends is small, the number of suspension support points is small, multiple routes are placed close together, a certain distance is maintained from other routes and obstacles, and when there are multiple pieces of equipment to connect at the endpoint and the route branches, the branching point is placed as close to the pieces of equipment as possible.

[0018] The invention described in claim 3 above lists specific items of the evaluation rules that the candidate routes should follow as much as possible.

[0019] As part of the present invention according to claim 4, the automatic routing method described in claim 1 is provided, wherein the evaluation function G is provided in multiple sets so that the user can select all or more of the following as items of importance to the user: the use of the target plant or building, the piping, duct, and wiring systems, the constituent materials, and the installation environment.

[0020] In the invention described in claim 4 above, for example, depending on the project, the user can select from multiple evaluation functions G according to the items they prioritize at that time, such as the use of the plant or building to be built, the type of piping system such as chilled and hot water piping, refrigerant piping, and water supply piping, the type of duct system such as supply air, exhaust air, and air conditioning air, the type of piping material such as carbon steel pipes, stainless steel pipes, and PVC pipes, the type of duct material such as galvanized steel sheets, stainless steel sheets, and galvalume steel sheets (registered trademark), or the installation environment such as outdoors, indoors, general environment, or special environment.

[0021] The present invention according to claim 5 is a method in which the termination condition for the iterative calculation in the fourth step and the method for selecting the automatically routed route are set to an upper limit on the number of iterations, and when the number of iterations reaches the upper limit, the candidate route from which the best evaluation index Y is obtained at that time is selected as the automatically routed route. The automatic routing method described in claim 1 is provided, which involves performing iterative calculations and, when the inter-cycle difference ΔY of the evaluation index Y with respect to the previous iterative calculation becomes less than or equal to a predetermined value set by the engineer or user as a convergence condition, adopting the candidate route for which the best evaluation index Y was obtained at that time as the automatically routed route.

[0022] The invention described in claim 5 above provides the termination conditions for the iterative calculation and the method for selecting the automatically routed route when performing the iterative calculation in the fourth step.

[0023] As part of the present invention according to claim 6, the automated routing method according to claim 1 is provided, which has a configuration and functions that allow the candidate route database to be classified, stored, and referenced according to one or more items selected from the use of the target plant or building, the piping, duct and wiring systems, the constituent materials, and the installation environment.

[0024] In the invention described in claim 6 above, since the rules for drawing differ depending on the use of the plant or building, the piping, duct, and wiring systems, the constituent materials, and the installation environment, the database for candidate routes is configured and has functions that allow it to be classified, stored, and referenced according to each of the above items.

[0025] As part of the present invention according to claim 7, an automatic routing method according to claim 1 is provided, in the fifth step, in which it is determined whether or not a route is similar to the user-modified route based on the criterion that the surrounding area between the two routes on the drawing is small.

[0026] In the invention described in claim 7 above, the criterion for determining whether the two routes, the database storage route and the user modification route, are similar is that the area enclosed between the two routes on the drawing is small.

[0027] As part of the present invention according to claim 8, the automatic routing method according to claim 1 is provided, which has a configuration and function that allows the modification history database to be classified, stored, and referenced according to one or more items selected from the use of the target plant or building, the piping, duct and wiring systems, the constituent materials, and the installation environment.

[0028] In the invention described in claim 8 above, in the sixth step, the modified or retained influence coefficient a i , the evaluation element function F i and the evaluation function G, user information of the operator who performed the operation, coordinate data of the user-modified route, coordinate data of the route most similar to the user-modified route among the candidate routes, and the influence coefficient a i , the evaluation element function F iIn addition to a configuration that stores the evaluation function G and the evaluation index Y in the revision history database, the system also includes a configuration and function that allows classification, storage, and referencing of the revision history database for one or more items selected from the target plant or building's purpose, piping / duct / wiring systems, constituent materials, and installation environment. This makes it easy for users to refer to the database information corresponding to various conditions for which they are creating drawings, enabling efficient utilization.

[0029] According to the present invention as of claim 9, in the fifth step, the influence coefficient a i As a way to correct this, evaluation index Y and evaluation element index x i If there is a positive correlation between the two, then the evaluation factor index x of the route adopted by automatic routing is i to x is , evaluation factor index x of the most similar candidate route i to x ik When we do this, we compare these and x is <x ik In this case, assuming that the rule being evaluated should take precedence over the current rule when performing automatic routing, the impact coefficient a i Increase x is >x ik In this case, assuming that the priority of the rule being evaluated can be lower than it is currently, the impact coefficient a i An automatic routing method according to claim 1 is provided, which performs an operation to reduce [something].

[0030] In the invention described in claim 9 above, the influence coefficient a i This document specifies the concrete methods for modifying these rules. This is expected to directly reflect the user's intended priority for each evaluation rule in the automated routing results.

[0031] The present invention according to claim 10 includes a first step of calculating candidate routes using a route search algorithm in a computer, The system checks whether the candidate route conforms to various mandatory rules that must be followed, and also checks to what extent the candidate route conforms to various evaluation rules that determine superiority or inferiority based on predetermined evaluation perspectives, using conditions predetermined by the engineer or user. or repetitive calculation formulas, arithmetic expression The evaluation element function F is composed of the above and takes coordinate data as input. i Evaluation element index x calculated from i The second step involves scoring using, The aforementioned evaluation factor index x i and the influence coefficient a that weights it i A third step involves calculating an overall evaluation index Y of the candidate routes using an evaluation function G that has the above as its components and determines the overall superiority or inferiority taking into account each of the aforementioned evaluation rules, The above steps 1 to 3 are repeated, and the coordinate data of the candidate route and the associated influence coefficient a are calculated with each iteration. i , the aforementioned evaluation element index x i A fourth step involves saving the evaluation index Y and the above evaluation index Y to a database for candidate routes. As a result of the iterative calculations performed in the fourth step above, the candidate route that is best based on the evaluation index Y is selected as the route to be automatically routed. Routes modified by the user as needed from the automatically routed routes are designated as user-modified routes. When a route stored in the aforementioned candidate route database is designated as a candidate database route, Among the candidate database routes, a search is performed to determine if there is a route similar to the user-modified route than the automatically routed route. If a similar route is found, the influence coefficient a is adjusted so that the automatically routed route approaches the user-modified route. i , the evaluation element function F i and a fifth step of modifying at least one of the evaluation functions G, The influence coefficient a that was modified or retained in the fifth step above. i , the evaluation element function F iand the evaluation function G, user information of the operator who performed the operation, coordinate data of the user-modified route, coordinate data of the route most similar to the user-modified route among the candidate routes, and the influence coefficient a i , the evaluation element function F i A program is provided for performing automatic routing, comprising a sixth step of saving the evaluation function G and the evaluation index Y in a database for modification history, and enabling referencing the data during subsequent routing. [Effects of the Invention]

[0032] As described in detail above, according to the present invention, manual correction work is minimized and highly accurate results can be obtained without the user having to manually change the settings of automatic routing or perform programming work. [Brief explanation of the drawing]

[0033] [Figure 1] This is a flowchart illustrating the implementation of a conventional, common automated routing method. [Figure 2] This flowchart shows the required rules and evaluation rules in conventional automated routing methods. [Figure 3] (A) is a 3D diagram showing the candidate routes output by automatic routing and the user-modified routes, and (B) is a 3D diagram showing the enclosed area, which is the difference between the two routes. [Figure 4] This is a flowchart illustrating the implementation of the automatic routing method according to the present invention. [Figure 5] This flowchart shows the required rules and evaluation rules in the automatic routing method according to the present invention. [Figure 6] This flowchart shows how to modify and save the values ​​for automatic routing based on user-modified routes, so that they can be referenced next time. [Figure 7] (A) and (B) are plan views showing examples of routes adopted by automatic routing and routes modified by the user. [Modes for carrying out the invention]

[0034] Embodiments of the present invention will be described in detail below with reference to the drawings.

[0035] [Traditional common automatic routing methods] First, we will explain conventional, general automatic routing methods based on Figures 1 and 2.

[0036] (Step 1) In the first step, candidate routes are calculated using a route search algorithm, as shown in Figure 1. The route search algorithm is a method that searches for routes through trial and error based on probability theory, such as a genetic algorithm, and conventionally known methods disclosed in, for example, Japanese Patent Publication No. 2002-288250 and Japanese Patent Publication No. 2002-351935 can be used without limitation.

[0037] (Step 2) In the second step, we check whether the candidate routes explored in the first step conform to various mandatory rules that must be followed, and to what extent the candidate routes conform to various evaluation rules that determine superiority or inferiority based on predetermined evaluation criteria. Engineer or An evaluation element function F, composed of user-defined conditional expressions and other elements, and taking coordinate data as input. i (i is the number of the rule being evaluated; the same applies hereinafter.) Evaluation element index x calculated from this. i The scoring is performed using [this method].

[0038] The aforementioned mandatory rules include one or more of the following: the target route (the candidate route to be evaluated for rule compliance) is connected without interruption between the specified start and end points; it does not interfere with other routes or obstacles (to maintain a certain distance); if the straight route between the start and end points is longer than the distance set by the user, it should be connected with a route that is as horizontal and vertical as possible (to avoid long-distance diagonal piping); and for routes requiring a horizontal gradient, such as drainage pipes, a gradient within a predetermined range should be ensured. Assuming that there are M mandatory rules, from Rule 1 to Rule M (where M is an integer of 1 or more), as shown in Figure 2, only routes that satisfy all of these rules from Rule 1 to Rule M can proceed to the next step. If even one route fails to comply, that candidate route is discarded, and the process returns to the first step above to calculate another candidate route.

[0039] The aforementioned evaluation rules include one or more of the following: the route length is short (to use the shortest possible route for piping), the number of bends is small (to reduce the number of joints), the number of suspension support points is small (to reduce the number of suspension support points), multiple routes are placed close together (to be supported by a common frame), they are placed at a certain distance from other routes and obstacles (to be supported by walls, etc.), and when there are multiple pieces of equipment to connect at the end point and the route branches, the branching point should be as close to the equipment as possible. Assuming that there are NM of the aforementioned evaluation rules, from rule M+1 to rule N (where M is an integer greater than or equal to 1 and N is an integer greater than M), as shown in Figure 2, the degree to which each rule is conformed is expressed by a score, and a higher score means a higher degree of conformance.

[0040] As an index representing the degree of conformity to the aforementioned evaluation rule, the evaluation element index x i The aforementioned evaluation element index x is used. i This is an evaluation element function F that takes the coordinate data of the candidate route and other data as input values. i The evaluation element function F is calculated from the above. i teeth, Engineer orThis is a calculation formula consisting of conditional expressions, repetitive calculation formulas, and arithmetic operation formulas predetermined by the user, and it outputs a numerical value indicating the degree of conformity to the rules being evaluated. (Step 3) The aforementioned evaluation factor index x i Candidate routes that have been scored by the evaluation factor index x are used in the third step. i and the influence coefficient a that weights it i The overall evaluation index Y of the candidate routes is calculated by an evaluation function G which uses the above evaluation rules as its components and determines the overall superiority or inferiority by taking each of the aforementioned evaluation rules into account (Y=G(f1(x1,a1),f2(x2,a2),...,f N-M (x N-M ,a N-M The evaluation function G is defined by the evaluation element index x i and the influence coefficient a i Functions such as linear and quadratic functions f i (x i ,a i It can be represented as ).

[0041] In step 3, once the overall evaluation index Y of the candidate route is calculated, the process returns to step 1 and similarly calculates the evaluation index Y for other candidate routes. By repeating this process, the best candidate route based on the evaluation index Y is output as the automatically routed route. In automatic routing methods such as genetic algorithms, when optimization is performed through iterative calculations, a termination condition for the iterative calculations is set. In Figure 1, an upper limit on the number of iterations is set, and when the number of iterations reaches the upper limit, the candidate route for which the best evaluation index Y was obtained at that point is output as the automatically routed route. The above calculation is repeated until all connection relationships between the start and end points set by the user are connected, and when all are connected, the plotting by automatic routing is terminated.

[0042] However, in actual field work, due to various circumstances, it is difficult to use the routes created by automated routing directly as construction drawings, and in most cases, users modify the routes on CAD drawings. Subsequently, as shown in Figure 1, users make modifications to the automated routing routes that have been output as needed, and the creation of construction drawings is completed. At this time, conventional automated routing methods did not have a function to feed back the user's drawing modifications to the automated routing method.

[0043] Thus, in existing automated routing methods using genetic algorithms (e.g., Japanese Patent Publication No. 7221434), the routes output by automated routing are determined by a pre-programmed evaluation element function F. i The evaluation element index x calculated by i Alternatively, methods have been devised to select routes based on whether the evaluation index Y calculated by the evaluation function G is superior. In contrast, the automatic routing method according to the present invention, described below, uses the content modified by the user on the route output by automatic routing as training data (correct answer), and uses "whether it is similar to the route modified by the user and ultimately adopted (correct answer)" as the criterion for judgment. As shown in Figure 3, it is characterized by operating while improving accuracy by incorporating a learning function, with the goal of minimizing the area enclosed by the route output by automatic routing (candidate route) and the route modified by the user (user modified route).

[0044] [Automatic routing method according to the present invention] Next, the automatic routing method according to the present invention will be described based on Figures 4, 5, and 6.

[0045] (Step 1) The first step is the same as the conventional automatic routing method described above, so we will omit the explanation.

[0046] (Step 2) The second step, similar to the conventional automated routing method described above, checks whether the candidate routes found in the first step conform to various mandatory rules that must be followed, and also checks to what extent the candidate routes conform to various evaluation rules that determine superiority or inferiority based on predetermined evaluation criteria. Engineer or User-defined conditional expression or repetitive calculation formulas, arithmetic expression The evaluation element function F is composed of the above and takes coordinate data as input. i Evaluation element index x calculated from i Scoring is performed using the evaluation element function F. i The various parameters used are not fixed values, but rather, as described later, are modified as needed during the iterative calculation process to improve the accuracy of the automatic routing results.

[0047] (Step 3) In the third step, similar to conventional automated routing methods, the evaluation factor index x i and the influence coefficient a that weights it i The overall evaluation index Y of the candidate routes is calculated by an evaluation function G which uses the above evaluation rules as its components and determines the overall superiority or inferiority by taking each of the aforementioned evaluation rules into account (Y=G(f1(x1,a1),f2(x2,a2),...,f N-M (x N-M ,a N-M The evaluation function G is, for example, an evaluation element index x i and the influence coefficient a i A linear function (f) whose constituent elements are the product of and i (x i a i Functions such as quadratic functions f i (x i ,a i It can be expressed as follows: that is, the evaluation element index x, which is the degree of conformity of each rule being evaluated. i Instead of simply adding up all the results, the overall evaluation index Y is calculated after adding weights that take into account the degree of influence of each evaluation rule on the overall evaluation. The influence coefficient a used for weighting is... iThis value is not a fixed number; as described later, it is modified as needed during the iterative calculation process to improve the accuracy of the automatic routing results.

[0048] In this invention, the evaluation function G is provided in multiple sets so that users can select all or more of the following items as important to them: the purpose of the target plant or building, the piping, duct, and wiring systems, the constituent materials, and the installation environment. For example, depending on the project, users may want to prioritize cost, energy efficiency after construction, or ease of construction and construction period. In such cases, users can appropriately select from the multiple evaluation functions G provided, depending on the purpose of the target plant or building, the type of piping system such as chilled / hot water piping, refrigerant piping, and water supply piping, the type of duct system such as supply air, exhaust air, and air conditioning air, the type of piping material such as carbon steel pipes, stainless steel pipes, and PVC pipes, the type of duct material such as galvanized steel sheets, stainless steel sheets, and galvalume steel sheets (registered trademark), or the installation environment such as outdoor, indoor, general environment, or special environment. When selecting from the above-mentioned group of items, users may select all of them, or they may select multiple items in appropriate combinations.

[0049] (Step 4) In the fourth step, steps 1 to 3 above are repeated, and the coordinate data of the candidate route and the associated influence coefficient a are calculated with each iteration. i , evaluation element index x i The evaluation index Y and the candidate route are stored in the database.

[0050] In the example shown in Figure 4, candidate routes similar to the user-modified diagram are reused from previous automated routing attempts to reduce the computer's computational load. Alternatively, the results of additional searches to find better candidate routes may also be used.

[0051] In the fourth step, the termination condition for the iterative calculation and the method for selecting the automatically routed route are as follows: In the example shown in Figure 4, an upper limit on the number of iterations is set, and when the number of iterations reaches the upper limit, the candidate route that yields the best evaluation index Y at that point is selected as the automatically routed route, and the iterative calculation is terminated. Alternatively, during the iterative calculation, when the inter-cycle difference ΔY of the evaluation index Y compared to the previous iterative calculation falls below a predetermined value set by the engineer or user as a convergence condition, the candidate route that yields the best evaluation index Y at that point is selected as the automatically routed route, and the iterative calculation is terminated.

[0052] Furthermore, in automatic routing methods, the design rules that should be prioritized differ depending on the purpose (e.g., whether it is in the ceiling or exposed, general or industrial use) and type (e.g., pressure piping for air conditioning, refrigerant piping, or drainage piping with a gradient) of the piping, ducts, and wiring. Therefore, it is preferable that the candidate route database be configured and functioned to classify, store, and reference data according to one or more items selected from the target plant or building's purpose, the piping, duct, and wiring system, the constituent materials, and the installation environment. By classifying, storing, and referencing the data stored in the candidate route database according to one or more items selected from the above group of items, the vast amount of data stored in the candidate route database can be effectively utilized, thereby improving the accuracy of automatic routing.

[0053] (Step 5) In the fifth step, as shown in Figures 4 to 6, feedback processing of the user's intended modifications is performed so that the intended modifications can be referenced during subsequent routing.

[0054] Specifically, as a result of the iterative calculations performed in step 4 above, the best candidate route based on the evaluation index Y is selected as the automatically routed route. Routes modified by the user as needed from these automatically routed routes are designated as user-modified routes. When a route stored in the aforementioned candidate route database is designated as a candidate database route, Among the candidate database routes, a search is performed to determine if there is a route similar to the user-modified route than the automatically routed route. If a similar route is found, the influence coefficient a is adjusted so that the automatically routed route approaches the user-modified route. i , the evaluation element function F i And at least one of the evaluation functions G has been modified.

[0055] Once the feedback process for this modification intention is complete, the user-modified route will be adopted as the final drawing. On the other hand, if there is no route similar to the user-modified route in the candidate database routes, the user-modified route will be adopted as the final drawing without any further feedback processing. Furthermore, if no modification by the user is required, the automatically routed route will be adopted as the final drawing.

[0056] The reason for the difference between automatically routed routes and user-modified routes is that the priority (weighting) of each evaluation rule differs between automatically routed and user-modified routes, or the evaluation factor index x i Evaluation element function F i This is thought to be due to an inappropriate calculation method for the evaluation function G. Therefore, as shown in Figures 4 to 6, the automatic routing method according to the present invention uses the modifications made by the user to the automatically routed routes as training data (correct answers), and uses the similarity of the user-modified routes (correct answers) as a criterion to determine whether they are similar to the user-modified routes (correct answers). The goal is to reduce the area enclosed by the routes that the user is likely to modify (see Figure 3(B)), and a function (feedback unit for modification intent) is added to feed back the user's modification intent into the evaluation elements of the automatic routing.

[0057] More specifically, as shown in FIG. 6, among the candidate database routes, a search is performed to determine whether there is a route similar to the user modification route among the automatic routing adoption routes. If there is a similar route, the influence coefficient a is adjusted so that the automatic routing adoption route approaches the searched user modification route. i , an operation of modifying at least one of the evaluation element function F i and the evaluation function G is performed. As a result, in the second and third steps, the evaluation element index x i is used for scoring and the comprehensive evaluation index Y reflects the user's intention in creating design and construction drawings. By repeating these operations for drawings and users at multiple sites, the influence coefficient a i , the evaluation element function F i or the evaluation function G is modified in a more reasonable form by collective intelligence, and a highly accurate automatic routing result close to the user's perception can be obtained. As a result, a highly accurate automatic routing adoption route can be obtained without the user manually changing the settings or programming the automatic routing.

[0058] (Step 6) In the sixth step, the influence coefficient a i , the evaluation element function F i and the evaluation function G, the user information of the operator who performed the operation, the coordinate data of the user modification route, and the coordinate data of the route most similar to the user modification route among the candidate routes, the influence coefficient a i , the evaluation element function F i , the evaluation function G and the evaluation index Y are stored in the database for modification history, enabling reference to each data during subsequent routing. The user information refers to information such as the name or title, identification number, etc. of the person who modified the candidate route output by the automatic routing, which enables the identifier of the modifier.

[0059] By saving and accumulating the calculation results up to step 5 above, as well as the user's modifications to the drawings and their user information, in a modification history database, and making this accessible in step 3, the convenience of the automatic routing function for users is improved, and new users can also share and utilize this information.

[0060] The aforementioned revision history database has a configuration and functionality that allows it to classify, store, and reference data by one or more items selected from the target plant or building's purpose, piping / duct / wiring systems, constituent materials, and installation environment. By classifying, storing, and referencing the data stored in the revision history database by one or more items selected from the above group of items, the vast amount of data stored in the revision history database can be effectively utilized, thereby improving the accuracy of automatic routing.

[0061] The sixth step, which involves saving the drawing modifications and the user information of the operator who made the modifications to a modification history database so that a modification history is kept, is expected to improve the convenience of modifying the automatic routing function. For example, based on a route modification made by a certain user, the influence coefficient a i When correcting the automatic routing adopted route a i (current a i (hereafter referred to as "a iA It states, ")) and the candidate route that is most similar to the route modified by the user a i (Example of correction a) i (hereafter referred to as "a iB It is written as follows: )) and Δa i There is a difference, that is, a iB =a iA +Δa i Let's look at cases where the relationship is as follows: Actual a i Correction amount Δa iB =Δa iTherefore, while automatic routing fully reflects the user's design intent from the previous modification, even if there is a history of modifications in the same case (same type and use of piping, ducts, etc.), that history will not be reflected in the next automatic routing. When a program implementing the present invention is used by multiple projects or an unspecified number of users, saving the modification history allows, for example, if there is a history of N modifications in the same case, Δa iB = 1 / N × Δa i By doing so, highly versatile routing can be performed as a result of collective intelligence. In addition, user information can be saved, and only route modifications made by specific highly skilled users can be reflected. i By modifying and deciding on these settings, it becomes possible to achieve near-perfect automated routing for specific users.

[0062] The information stored in the aforementioned revision history database may include the following items regarding the current settings and past history. One or more of these items can be arbitrarily selected and saved.

[0063] (1) Current setting Building use • Types of piping, ducts, and wiring • Other Attribute Information for Piping, Ducts, and Wiring • Influence coefficient a i • Evaluation element function F i • Evaluation function G

[0064] (2) Past history • User information that has been modified • Coordinate data of user-modified routes • The building use, type of piping / ducting / wiring, other attribute information of piping / ducting / wiring, and influence coefficient a of the route that is most similar to the user-modified route among the candidate routes. i , evaluation element function F i Evaluation function G

[0065] [Specific examples] The effects of the present invention will be explained by giving a simple example, as shown in Figure 7, of creating a first route connecting starting point 1 to ending point 1 and a second route connecting starting point 2 to ending point 2.

[0066] The following mandatory rules and evaluation rules will be used as design rules when plotting each route.

[0067] • Required rules (1) The specified start and end points must be connected without interruption. (2) Ensure that it does not interfere with other routes or obstacles (maintain a certain distance). (3) If the straight route exceeds a specified distance, connect using a horizontal or vertical route whenever possible (avoid long-distance diagonal piping).

[0068] • Evaluation Rules (4) The route length should be short (pipe through the shortest possible route). (5) The number of bends (joints) is small. (6) Arranging multiple routes close together (common support structure) (7) Position it at a certain distance from other routes and obstacles (walls) (provide support). (8) When there are multiple pieces of equipment to connect at the terminal and the route branches, the branching point should be as close to the equipment as possible.

[0069] Figure 7(A) shows the route obtained using a conventional automated routing method based on the design rules described above, shown as a dotted line, and the route modified by the user, shown as a solid line. The difference (enclosed area) between these routes is the gray-filled area for the first route and the hatched area for the second route.

[0070] Figure 7(B) shows the route obtained using the automatic routing method according to the present invention, based on the design rules described above, as shown by the dashed line, and the route modified by the user as shown by the solid line.

[0071] For both the first and second routes, the results obtained using the automatic routing method according to the present invention (B) show a smaller difference from the user-modified route and require less user modification than the results obtained using the conventional automatic routing method (A). Therefore, routes closer to the user's design intent can be created.

[0072] Comparing each route with the design rules, we can analyze that in the conventional, general automated routing method shown by the dotted line, (4) and (5) are prioritized, while in the user-modified route, (6), (7), and (8) are prioritized. Evaluation index Y and evaluation element index x i If there is a positive correlation between the two, then the evaluation element index x i The larger x is, the greater its impact on the evaluation index Y, therefore the x adopted in automatic routing i to x iS , the candidate route x that is most similar to the user-modified route among the candidate routes i to x iK So, x 4S >x 4K , x 5S >x 5K , x 6S <x 6K , x 7S <x 7K , x 8S <x 8K It is thought that they are in the following relationship. Furthermore, evaluation index Y and evaluation element index x i If there is a negative correlation between the two, the signs will be reversed.

[0073] In the automated routing method according to the present invention, for example, the influence coefficients a4 and a5 are reduced to smaller values ​​than their current values, while the influence coefficients a6, a7, and a8 are increased to larger values. As a result, in the next automated routing, the influence of the evaluation target rules (4) and (5) will be relatively lower than the previous time, while the influence of the evaluation target rules (6), (7), and (8) will be relatively higher than the previous time. Consequently, the amount of manual correction required by the user in the next round can be further reduced, leading to labor savings in the creation of design and construction drawings.

[0074] [Other examples of forms] Furthermore, in the fifth step described above, if there are candidate routes that are more similar to the user-modified routes than the automatically routed routes, one example of how to modify the evaluation function G is to modify the evaluation function G, which has a neural network structure, by modifying the weight coefficients of the neural network so that it approaches the training data.

Claims

1. The first step is to calculate candidate routes using a route search algorithm, The system checks whether the candidate route conforms to various mandatory rules that must be followed, and also checks to what extent the candidate route conforms to various evaluation target rules that determine superiority or inferiority based on predetermined evaluation perspectives. This is done using an evaluation element function F, which consists of conditional expressions, repetitive calculation formulas, and arithmetic operations predetermined by the engineer or user, and takes coordinate data as input. ｉ Evaluation factor index x calculated from ｉ The second step involves assigning scores using, The aforementioned evaluation factor index x ｉ and the influence coefficient a that weights it ｉ The third step involves calculating an overall evaluation index Y of the candidate routes using an evaluation function G that has the above as its components and determines the overall superiority or inferiority taking into account each of the aforementioned evaluation rules, The above steps 1 to 3 are repeated, and the coordinate data of the candidate route and the associated influence coefficient a are calculated with each iteration. ｉ , the aforementioned evaluation element index x ｉ A fourth step involves saving the evaluation index Y and the above evaluation index Y to a database for candidate routes. As a result of the iterative calculations performed in the fourth step above, the candidate route that is best based on the evaluation index Y is selected as the route to be automatically routed. Routes modified by the user as needed from the automatically routed routes are designated as user-modified routes. When a route stored in the aforementioned candidate route database is designated as a candidate database route, Among the candidate database routes, a search is performed to determine if there is a route similar to the user-modified route than the automatically routed route. If a similar route is found, the influence coefficient a is adjusted so that the automatically routed route approaches the user-modified route. ｉ , the evaluation element function F ｉ And a fifth step of modifying at least one of the evaluation functions G, The influence coefficient a corrected or maintained in the fifth step ｉ , the evaluation element function F ｉ and the evaluation function G, user information of the operator who performed the operation, coordinate data of the user correction route, and coordinate data of the route in the candidate routes that is most similar to the user correction route, the influence coefficient a ｉ , the evaluation element function F ｉ , the evaluation function G and the evaluation index Y are stored in a database for correction history, and a sixth step that enables reference to each data during subsequent routing, and an automatic routing method characterized by being composed of the above.

2. The automatic routing method according to claim 1, wherein the aforementioned required rules include one or more of the following: the target route is connected without interruption between the specified start point and end point; it does not interfere with other routes or obstacles; the route between the start point and end point is connected as horizontally and vertically as possible; and in routes where a horizontal gradient is required for a horizontal pipe, a gradient within a predetermined range is ensured.

3. The automatic routing method according to claim 1, wherein the evaluation rules include one or more of the following: the route length is short, the number of bends is small, the number of suspension support points is small, multiple routes are placed close together, a certain distance is maintained from other routes and obstacles, and when there are multiple pieces of equipment to connect at the endpoint and the route branches, the branching point is placed as close to the pieces of equipment as possible.

4. The automatic routing method according to claim 1, wherein the evaluation function G is provided in multiple sets so that the user can select all or more of the following items as important to the user: the purpose of the target plant or building, the piping, duct and wiring system, the constituent materials, and the installation environment.

5. The termination condition for the iterative calculation in the fourth step and the method for selecting the automatically routed route are as follows: A maximum number of iterations is set, and when the number of iterations reaches the maximum number, the candidate route from which the best evaluation index Y is obtained at that time is selected as the automatically routed route. The automatic routing method according to claim 1, wherein, during the iterative calculation, when the intercycle difference ΔY of the evaluation index Y with respect to the previous iterative calculation becomes less than or equal to a predetermined value set by the engineer or user as a convergence condition, the candidate route from which the best evaluation index Y is obtained at that time is adopted as the automatically routed route.

6. The automated routing method according to claim 1, wherein the candidate route database has a configuration and functions that allow it to classify, store, and reference one or more items selected from the use of the target plant or building, the piping, duct and wiring systems, the constituent materials, and the installation environment.

7. The automatic routing method according to claim 1, wherein in the fifth step, whether or not a route is similar to the user-modified route is determined based on the fact that the surrounding area between the two routes on the drawing is small.

8. The automatic routing method according to claim 1, wherein the database for modification history has a configuration and functions that allow it to classify, store, and refer to one or more items selected from the use of the target plant or building, the piping, duct and wiring systems, the constituent materials, and the installation environment.

9. In the fifth step, the influence coefficient a ｉ As a method for correcting this, evaluation index Y and evaluation element index x ｉ If there is a positive correlation between the two, then the evaluation factor index x of the route adopted by automatic routing. ｉ to x ｉｓ , evaluation factor index x of the most similar candidate route ｉ to x ｉｋ When we do this, we compare these and x ｉｓ <x ｉｋ In this case, assuming that the rule being evaluated should take precedence over the current rule when performing automatic routing, the impact coefficient a ｉ Increase x ｉｓ > x ｉｋ In this case, assuming that the priority of the rule being evaluated can be lower than it is currently, the impact coefficient a ｉ The automatic routing method according to claim 1, which performs an operation to reduce.

10. In a computer, the first step is to calculate candidate routes using a route search algorithm, The system checks whether the candidate route conforms to various mandatory rules that must be followed, and also checks to what extent the candidate route conforms to various evaluation target rules that determine superiority or inferiority based on predetermined evaluation perspectives. This is done using an evaluation element function F, which consists of conditional expressions, repetitive calculation formulas, and arithmetic operations predetermined by the engineer or user, and takes coordinate data as input. ｉ Evaluation factor index x calculated from ｉ The second step involves assigning scores using, The aforementioned evaluation factor index x ｉ and the influence coefficient a that weights it ｉ The third step involves calculating an overall evaluation index Y of the candidate routes using an evaluation function G that has the above as its components and determines the overall superiority or inferiority taking into account each of the aforementioned evaluation rules, The above steps 1 to 3 are repeated, and the coordinate data of the candidate route and the associated influence coefficient a are calculated with each iteration. ｉ , the aforementioned evaluation element index x ｉ A fourth step involves saving the evaluation index Y and the above evaluation index Y to a database for candidate routes. As a result of the iterative calculations performed in the fourth step above, the candidate route that is best based on the evaluation index Y is selected as the route to be automatically routed. Routes modified by the user as needed from the automatically routed routes are designated as user-modified routes. When a route stored in the aforementioned candidate route database is designated as a candidate database route, Among the candidate database routes, a search is performed to determine if there is a route similar to the user-modified route than the automatically routed route. If a similar route is found, the influence coefficient a is adjusted so that the automatically routed route approaches the user-modified route. ｉ , the evaluation element function F ｉ And a fifth step of modifying at least one of the evaluation functions G, The influence coefficient a that was modified or retained in the fifth step above. ｉ , the evaluation element function F ｉ and the evaluation function G, user information of the operator who performed the operation, coordinate data of the user-modified route, coordinate data of the route most similar to the user-modified route among the candidate routes, and the influence coefficient a ｉ , the evaluation element function F ｉ A program for performing automatic routing, comprising a sixth step of saving the evaluation function G and the evaluation index Y in a database for modification history, and enabling referencing of the respective data during subsequent routing.

Images