Automatic splitting method for ship pipe assembly

Abstract

The invention discloses a ship pipe assembly automatic splitting method, which comprises the following steps of 1, constructing a pipe assembly drawing structure model, and performing drawing structure representation on a finished pipe model; 2, splitting the graph structure model of the pipe assemblage through an improved split hierarchical clustering algorithm; 3, evaluating the splitting and clustering result in the step 2, and when the evaluation result does not meet the requirement, returning to the step 2 for re-splitting until the evaluation result meets the requirement; and step 4, encoding the split assemblage to complete automatic splitting of the pipe assemblage. According to the method, the graph structure is combined with the clustering algorithm, so that intelligent splitting of the pipe assemblage is realized, dependence on artificial experience is avoided, and the splitting efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of ship pipe processing, and particularly to an automatic splitting method for ship pipe assemblies. Background Art

[0002] The ship piping system is an important part of the ship power plant. It connects the main and auxiliary machinery and related equipment to ensure the normal navigation, berthing of the ship, and the needs of crew and passengers. Pipe manufacturing is an important link in shipbuilding, including multiple processes such as pipe material cutting, bending, welding, and surface treatment, accounting for about 15% to 30% of the shipbuilding cycle. The pipe manufacturing of ships has always been an important part of the transformation and modeling work in the shipbuilding industry, and the progress of pipe manufacturing will directly affect the overall shipbuilding progress.

[0003] During the shipbuilding process, it is necessary to reasonably arrange the progress of pipe manufacturing to ensure its coordinated progress with other links to improve the shipbuilding efficiency. Flow-line manufacturing can improve the pipe manufacturing efficiency, and the adoption of flow-line manufacturing requires splitting and assembling. In the traditional pipe processing and manufacturing process, manual experience is usually relied on for splitting and determining the manufacturing flow, resulting in low efficiency and easy errors. With the development of technology, the application of automation technology in the manufacturing industry is becoming more and more extensive, but in the field of pipe manufacturing, there is a lack of an efficient and accurate method for automatically splitting the manufacturing flow.

[0004] The existing production methods of pipes are mainly divided into two types: one is based on group technology, adopting the "integral assembly and welding" mode, where overall assembly is carried out first and then overall welding. This method has a relatively simple workshop layout, but the production efficiency is low. The second is the "subsequent assembly and welding" mode, where the pipes are first split into multiple small assemblies for assembly and welding, and then overall assembly and welding are carried out. This method has a slightly more complex workshop layout, but the manufacturing efficiency is higher than the first one.

[0005] Currently, most shipbuilding enterprises adopting the "subsequent assembly and welding" mode usually rely on manual experience for splitting and determining the manufacturing flow, resulting in low efficiency and easy errors. Summary of the Invention

[0006] In view of the deficiencies in the prior art, the present invention provides an automatic splitting method for ship pipe assemblies to solve the technical problems of low efficiency and high error rate in manual splitting in the prior art.

[0007] The present invention provides an automatic splitting method for ship pipe assemblies, including the following steps:

[0008] Step 1: Construct a graph structure model of the pipe assembly and represent the finished pipe model in a graph structure;

[0009] Step 2: Split the graph structure model of the pipe assembly by an improved split hierarchical clustering algorithm;

[0010] Step 3: Evaluate the splitting and clustering results of Step 2. When the evaluation results do not meet the requirements, return to Step 2 to perform splitting again until the evaluation results meet the requirements;

[0011] Step 4: Encode the split assemblies to complete the automatic splitting of pipe assemblies.

[0012] Further, in the said Step 1, the specific method for constructing the pipe assembly graph structure model includes:

[0013] Define each pipe fitting as a node in the graph structure and distinguish the nodes by labels;

[0014] Construct the edges between the nodes according to the connection relationships between the pipe fittings, and distinguish the edges according to the attributes of the connection methods.

[0015] Further, the specific method for distinguishing the edges according to the attributes of the connection methods is:

[0016] Use a weight value to represent the strength attribute or importance attribute of the connection method.

[0017] Further, the specific method of the said Step 2 is:

[0018] Step 21: Extract the node feature information, construct the feature vectors and generate the data matrix;

[0019] Step 22: Classify all nodes into a single initial cluster;

[0020] Step 23: Perform a layer-by-layer judgment and splitting process on the pipe assembly graph structure, specifically:

[0021] Split the top-level node from the initial cluster, and perform the following judgment on the subsequent layer nodes:

[0022] When the node connected to the vertex has no next-layer nodes, split the current node from the initial cluster and merge it with the vertex into a cluster;

[0023] When the current node is a node representing an elbow and the node connected to it is a pipe, split the current node, the node representing the pipe connected to it, and the subsequent nodes representing flanges connected to it from the initial cluster, and merge the split nodes into a new cluster;

[0024] When the current node is a node representing a flange, split the current node and the node connected to it from the initial cluster, and merge the split nodes into a new cluster;

[0025] When the current node has several downward branches, split the current node and the branches from the initial cluster and perform a layer-by-layer judgment and splitting process separately;

[0026] Step 24: Stop splitting when there are no lower-level nodes in all new clusters.

[0027] Further, in the above step 3, the method for evaluating the split clustering result is as follows:

[0028] Obtain the silhouette coefficient for each node in the pipe assembly, calculate the average value of the silhouette coefficients of all nodes, and evaluate the split clustering result based on the average value of the silhouette coefficients.

[0029] Further, the calculation formula for the silhouette coefficient is as follows:

[0030]

[0031] In the formula, is the average distance from node P i to other nodes within its affiliated pipe assembly; is the average distance from node P i to nodes within the adjacent pipe assembly.

[0032] Further, in the above step 4, the method for encoding the split assemblies is as follows:

[0033] Construct a three-digit code, where the first digit represents the pipe fitting type in the assembly, the second digit represents the number of accessories, and the third digit represents the manufacturing step.

[0034] Advantages of the present invention:

[0035] By combining the graph structure with the clustering algorithm, the present invention realizes intelligent splitting of pipe assemblies, avoids dependence on manual experience, and improves the splitting efficiency. The present invention quantifies and evaluates the splitting result through the silhouette coefficient to ensure the scientificity and rationality of the splitting, and at the same time improves the accuracy of the splitting. The standardized coding system of the present invention simplifies the process flow, reduces the misjudgment rate during on-site construction, and can shorten the overall construction period. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings. The drawings are schematic and should not be construed as imposing any limitation on the present invention. In the drawings:

[0037] Figure 1 is a flowchart of a specific embodiment of the present invention;

[0038] Figure 2 is a schematic diagram of a finished pipe model used in a specific embodiment of the present invention;

[0039] Figure 3 is a schematic diagram of the graph structure of the pipe assembly in a specific embodiment of the present invention;

[0040] Figure 4 is a schematic diagram of the split assembly in a specific embodiment of the present invention. Detailed implementation manners

[0041] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0042] The present invention will be further clarified below with reference to specific embodiments. Those skilled in the art should understand that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. Modifications to various equivalent forms of the present invention all fall within the scope defined by the appended claims of this application.

[0043] The present invention provides an automatic splitting method for ship pipe assemblies, including the following steps:

[0044] Step 1: Construct a graph structure model of the pipe assembly drawing and perform a graph structure representation on the finished pipe model. The specific method for constructing the graph structure model of the pipe assembly drawing includes:

[0045] Define each pipe fitting as a node in the graph structure and distinguish the nodes through labels;

[0046] Construct edges between the nodes according to the connection relationship between the pipe fittings, and distinguish the edges according to the attributes of the connection methods.

[0047] The finished pipe model is as Figure 2 shown. The finished pipe model includes the geometric information of the pipe, such as length, diameter, bending angle, etc.; the connection information of the pipe, such as the types and positions of connectors such as flanges, elbows, tees, etc.; the material information of the pipe, such as the material of the pipe material.

[0048] Analyze the finished pipe model, identify the pipe parts therein, including main pipes, branch pipes, valve parts, elbows, etc., and define each pipe part as a node in the graph structure. As Figure 3 shown, use letters A to O to represent Figure 2The pipe parts included in the shown finished pipe model, and add a text attribute to each node as a label to distinguish the node types. For example, nodes B, D, F, I, L, O, and N represent flanges, and add the label "flange"; node G represents an elbow, and add the label "pipe elbow"; nodes A and H represent straight pipes, and add the label "straight pipe"; if the node represents a bent pipe, add the label "bent pipe"; nodes C, J, K, and M represent straight branch pipes, and add the label "straight branch pipe", and node E represents a bent branch pipe, and add the label "bent branch pipe".

[0049] Analyze the connection relationships between pipe parts, such as welding, flange connection, threaded connection, etc., and establish edges between the corresponding nodes according to the connection relationships. For different types of connections, they can be distinguished by the attributes of the edges. For example, different weight values are used to represent the strength or importance of the connections. For welded connections, the weight is 1; for flange connections, the weight is 2; for threaded connections, the weight is 3. Let the connection weight between nodes i and j be w ij , then the formula for defining the weight of the edge is:

[0050]

[0051] For example Figure 3 in, the number "1" is marked on the connection line between nodes, indicating that the nodes are welded.

[0052] Step 2: Split the graph structure model of the pipe erection by an improved split hierarchical clustering algorithm;

[0053] The specific method is as follows:

[0054] Step 21: Extract the node feature information, construct the feature vectors and generate the data matrix;

[0055] For Figure 3 the 15 nodes in, extract the features of each node. For example, the hierarchical level l of node A A = 1, the number of connected pipes n pA = 4, the spatial position is (x A , y A , z A ), the pipe diameter d A , and if there is a branch, whether the node has a branch b A = 1. Combine these features into a feature vector Construct a 15 * 5 data matrix;

[0056] Step 22: Classify all nodes into a single initial cluster;

[0057] Place 15 nodes in a single cluster, and this initial cluster contains all node information of the entire pipe erection system from the top - level node to the bottom - level node;

[0058] Step 23: Perform a layer - by - layer judgment and splitting process on the pipe erection diagram structure, specifically as follows:

[0059] Split the top - level node from the initial cluster, and perform the following judgment on the subsequent - layer nodes:

[0060] When the node connected to the vertex has no next - layer node, split the current node from the initial cluster and merge it with the vertex into a cluster;

[0061] When the current node is a node representing an elbow and the node connected to it is a pipe, split the current node, the node representing the pipe connected to it, and the subsequent - connected node representing a flange from the initial cluster, and merge the split - out nodes into a new cluster;

[0062] When the current node is a node representing a flange, split the current node and the node connected to it from the initial cluster, and merge the split - out nodes into a new cluster;

[0063] When the current node has several downward branches, split the current node and its branches from the initial cluster and perform a layer - by - layer judgment and splitting process separately.

[0064] Take Figure 3 as an example, the layer - by - layer judgment and splitting process is as follows: Node A is the vertex, split the top - level node A. For the subsequent - layer node judgment and splitting process: Node B has no next - layer node, merge B without a second - layer node and A into a cluster, named C1; Node C, E, G, J and their lower - layer nodes are grouped into another cluster. For the cluster composed of Node C, E, G, J and their lower - layer nodes, first apply the principle of merging flange nodes with pipe nodes and elbow nodes with pipe and flange nodes. Merge C and D into cluster C2, merge E and F into cluster C3; merge G, H, I into cluster C4; for the unmerged nodes: J and its lower - layer nodes, etc., process them separately according to the layer - by - layer judgment and splitting process. Split the first - layer node O without a second - layer node and the vertex J to form an erection cluster C5; merge K and L into cluster C6; merge M and N into cluster C7.

[0065] Step 24: When there are leaf nodes (i.e., no lower - layer nodes) in clusters C1 to C7, stop splitting, and the splitting result is as Figure 4 shown.

[0066] Step 3: Evaluate the splitting and clustering results of Step 2. When the evaluation result does not meet the requirements, return to Step 2 to re - perform the splitting until the evaluation result meets the requirements;

[0067] The method for evaluating the split clustering results is as follows:

[0068] For each node in the pipe erection, obtain the silhouette coefficient, calculate the average value of the silhouette coefficients of all nodes, and evaluate the split clustering results through the average value of the silhouette coefficients. The calculation formula for the silhouette coefficient is as follows:

[0069]

[0070] In the formula, is the average distance from node P i to other nodes within its affiliated pipe erection. Suppose there are N nodes in the cluster, and each other node within the cluster to which node P i belongs is P j , then:

[0071]

[0072] is the average distance from node P i to nodes within the adjacent pipe erection. Suppose there are M nodes in the adjacent cluster. For each node P k within the adjacent cluster, calculate the distance between node P i and node P k . Add up all these distances and divide by the number of nodes in the adjacent cluster, then:

[0073]

[0074] Finally, calculate the average value for all nodes to obtain the silhouette coefficient of the entire clustering result, that is

[0075] Step 4: Encode the split erections to complete the automatic split of the pipe erections.

[0076] For the convenience of on-site construction personnel to quickly identify and accelerate the transfer of parts between workstations, encode the split pipe erections. The pipe erection code consists of three digits. Among them, the first digit represents the pipe fitting type in the erection, the second digit represents the number of accessories, and the third digit represents the manufacturing step, as shown in Table 1 below.

[0077] Assembly Code: Letter + X (Number of Accessories) + X (Manufacturing Steps) Letter Explanation: A Short Straight Pipe AL Long Straight Pipe B Bent Pipe M Main Pipe BM Bent Main Pipe R Branch Pipe BR Bent Branch Pipe E With Elbow and Reducer D Large Group Explanation of the Number of Accessories: 0 Pipe Section with 0 Accessories 1 Pipe Section with 1 Accessory 2 Pipe Section with 2 Accessories Explanation of Manufacturing Steps: 0 Completed 1 Transfer to Small Group / Elbow Station 2 Transfer to Large Group Station

[0078] Table 1

[0079] On-site construction personnel can quickly implement the subsequent processing process based on the encoding of the split erections.

[0080] Although embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations fall within the scope defined by the appended claims.

Claims

1. An automatic splitting method for ship pipe assemblies, characterized in that, It includes the following steps: Step 1: Construct a structural model of the pipe erection drawing and represent the finished pipe model in a graph structure; Step 2: Split the graph structure model of the pipe erection by an improved split hierarchical clustering algorithm; Step 3: Evaluate the split clustering result of Step 2. When the evaluation result does not meet the requirements, return to Step 2 to re - split until the evaluation result meets the requirements; Step 4: Encode the split erections to complete the automatic splitting of the pipe erection.

2. The automatic disassembly method of the ship pipe assembly according to claim 1, characterized in that In the above - mentioned Step 1, the specific method for constructing the graph structure model of the pipe erection includes: Define each pipe fitting as a node in the graph structure and distinguish the nodes by labels; Construct the edges between the nodes according to the connection relationship between the pipe fittings, and distinguish the edges according to the attributes of the connection method.

3. The automatic splitting method of the ship pipe assembly according to claim 2, characterized in that, The specific method for distinguishing the edges according to the attributes of the connection method is: Use a weight value to represent the strength attribute or importance attribute of the connection method.

4. The automatic splitting method of ship pipe assemblies as described in claim 1, characterized in that, The specific method of the above - mentioned Step 2 is: Step 21: Extract node feature information, construct a feature vector and generate a data matrix; Step 22: Classify all nodes into a single initial cluster; Step 23: Perform a layer - by - layer judgment and splitting process on the graph structure of the pipe erection. Specifically: Split the top - layer node from the initial cluster, and perform the following judgment on the subsequent - layer nodes: When the node connected to the vertex has no next - layer node, split the current node from the initial cluster and merge it with the vertex into a cluster; When the current node is a node representing an elbow and the node connected to it is a pipe, split the current node, the node representing the pipe connected to it, and the subsequent - connected node representing a flange from the initial cluster, and merge the split nodes into a new cluster; When the current node is a node representing a flange, split the current node and the node connected to it from the initial cluster, and merge the split nodes into a new cluster; When the current node has several downward branches, split the current node and its branches from the initial cluster and separately perform a layer - by - layer judgment and splitting process; Step 24: Stop splitting when there are no lower - layer nodes in all new clusters.

5. The automatic disassembly method of the ship pipe assembly as described in claim 1, characterized in that, In the above - mentioned Step 3, the method for evaluating the split clustering result is: Obtain the silhouette coefficient for each node in the pipe erection, calculate the average value of the silhouette coefficients of all nodes, and evaluate the split clustering result through the average value of the silhouette coefficients.

6. The automatic disassembly method of the ship pipe assembly as claimed in claim 5, wherein The calculation formula of the silhouette coefficient is: In the formula, is the average distance from node P i to other nodes within its affiliated pipe assembly; is the average distance from node P i to nodes within an adjacent pipe assembly.

7. The automatic splitting method of the ship pipe assembly as described in claim 1, characterized in that, In the above - mentioned Step 4, the encoding method for the split erections is: Construct a three - digit code. Among them, the first digit represents the type of pipe fitting in the erection, the second digit represents the number of accessories, and the third digit represents the manufacturing step.

Images