Intelligent quotation and inquiry management system and method for pressure vessels

Abstract

The application discloses a kind of pressure vessel intelligent quotation and inquiry management system and method, comprising: weight estimation module, based on equipment overall parameter and historical template fingerprint matching fast generation estimation weight table;Quotation details generation module, based on parameter-part mapping rule library automatic disassembly generates BOM table;Intelligent classification extraction module, based on multi-keyword recognition and regular expression analysis classifies parts into head, cylinder, flange, gasket, nozzle and other categories;Inquiry sheet generation module, reconstruct field removes price sensitive information, matches supplier database to generate directional inquiry sheet.The application realizes the full automation from weight estimation to supplier inquiry, shortens the quotation-inquiry process of traditional 4-6 hours to minute level, while ensuring data consistency and business secret safety.

Description

Technical Field

[0001] This invention relates to the field of quotation management and supply chain collaboration technology in the pressure vessel manufacturing industry, and in particular to an intelligent quotation and inquiry management system and method for pressure vessels. Background Technology

[0002] Pressure vessels are core equipment in industries such as petroleum, chemical, and energy, and their pricing and procurement processes involve multiple complex steps.

[0003] Traditional business processes mainly rely on manual labor, thus presenting the following technical challenges:

[0004] The first level (weight estimation) pain point: Traditional weight estimation relies on engineers manually checking historical data, calculating material costs, and calculating processing time. It takes 2-5 working days to quote a single piece of equipment. Historical quotation data is stored in a scattered manner, making it difficult to quickly match similar equipment for reference quotations.

[0005] The second level (detailed quotation) pain points: The detailed BOM (Bill of Materials) generated from the overall parameters of the equipment relies on the personal experience of the engineer. Novices find it difficult to accurately determine what specifications of pipes, what pressure rating of flanges, and what gasket standard should be configured for DN3600 equipment; the price coefficients of parts (head 1.40, shell 1.05, pipe 2.00, etc.) lack transparent calculation logic, and manual maintenance is prone to errors.

[0006] Pain points at the third level (inquiry extraction): The high omission rate of manually screening purchased parts (gaskets, flanges, end caps, etc.) from the BOM; Excel macros can only perform simple keyword searches and cannot identify specification differences. Copying entire rows will also bring out price-sensitive information, making it unsuitable for directly sending out inquiries; There is a lack of automatic supplier matching, and manual allocation of recipients is required. Summary of the Invention

[0007] This invention provides an intelligent quotation and inquiry management system and method for pressure vessels, which automates the entire process from weight estimation to supplier inquiry, solving the technical problems of low efficiency and poor consistency in traditional processes.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0009] A method for intelligent quotation and inquiry management of pressure vessels is provided, which includes the following steps:

[0010] Intelligent generation of S1 weight estimation table

[0011] S1.1 Obtain or extract the basic parameters of the equipment to be quoted. The basic parameters include one or more of the following: equipment tag number, equipment name, design pressure, design temperature, medium characteristics, material grade, nominal diameter, tangential length, or design standard.

[0012] The sources of the basic parameters include images in different formats such as email, PDF, and JPG, or documents in formats such as Word and Excel.

[0013] S1.2 Construct the current operating condition feature vector V=[P,T,M,V,S,C,G] of the equipment to be queried based on the acquired or extracted basic parameters for easy querying. Where P is the design pressure, T is the design temperature, M is the medium type, V is the volume, S is the design standard, C is the material type, and G is the special process requirements.

[0014] S1.3 Calculate the similarity between the current working condition feature vector and each record in the historical weight estimation table based on weighted Euclidean distance, and filter out historical records that match the working condition feature vector according to the preset similarity threshold, and dynamically generate a benchmark weight estimation table.

[0015] The baseline weight estimation table is a comprehensive table with weight data (unit weight) as the core, and also includes multi-dimensional information such as part configuration, material standards, specifications, and price coefficients. It serves as the data foundation for subsequent BOM breakdown and quotation generation.

[0016] If multiple historical records are selected, they can be sorted in descending order of similarity, and then the most similar historical records can be selected based on a preset quantity threshold.

[0017] S1.4 Adjust the baseline weight estimation table according to the real-time correction factors to generate the final weight estimation table. The real-time correction factors include the material price index, exchange rate fluctuation coefficient, and capacity load coefficient.

[0018] Automatic decomposition of S2 quotation details

[0019] S2.1 Based on the equipment type-part configuration knowledge graph, automatically extract and generate a detailed Bill of Materials (BOM) from the final weight estimation table;

[0020] The knowledge graph includes: horizontal storage tanks mapping double heads, double saddles, manholes, and connecting pipe assemblies; vertical reactors mapping double heads, skirts, stirring interfaces, and multi-specification connecting pipe assemblies; and heat exchangers mapping shell-side shells, tube boxes, tube sheets, and heat exchange tube bundles.

[0021] The BOM table includes multiple category fields, and the disassembled parts or parameter information is filled into the corresponding category fields. The category fields include part name field, specification field, material field, quantity field, unit weight field, coefficient field, unit price field, subtotal field, and remarks field.

[0022] S2.2 Based on the parameter-part mapping rule base, automatically determine the material standards and size specifications of each part in the BOM table;

[0023] The parameter-part mapping rule library includes design pressure-flange rating rules, medium characteristics-gasket type mapping rules, and diameter-pipe specification mapping rules.

[0024] S2.3 Calculate the price coefficient of each part in the BOM based on the comprehensive coefficient decomposition model of the parts. The decomposition model includes: material forming coefficient, processing difficulty coefficient, quality grade coefficient, and batch discount coefficient.

[0025] Price coefficients are used to calculate internal quotation costs (subtotals), which directly affect the calculation of quotation amounts. The calculated subtotal data needs to be recorded in the BOM table. Subtotals refer to quotations provided to customers and are considered sensitive information. They will only be reflected in the complete BOM table and do not need to be sent to suppliers. Therefore, they need to be removed from the inquiry form.

[0026] The formula for calculating the subtotal is: Subtotal = Unit weight × Coefficient × Unit price. The higher the coefficient, the greater the proportion of the processing cost of the part (e.g., 1.40 for the end cap includes mold amortization + 15% forming loss, and 2.00 for the nozzle includes forging procurement + machining cost).

[0027] S3 Product Category Intelligent Extraction and Inquiry Form Generation Steps

[0028] S3.1 Based on a multi-keyword parallel recognition algorithm, the parts in the BOM are classified into preset part categories;

[0029] The parts categories include: heads, cylinders, flanges, gaskets, pipes, and support structures.

[0030] S3.2 uses regular expressions to perform structured parsing of the specifications field in the BOM table and extracts part parameters, including nominal diameter, wall thickness, length, pressure rating, and standard number.

[0031] S3.3 Based on the part category, match the supplier database to determine the default supplier and dedicated inquiry template for each part;

[0032] S3.4 Based on the inherent category characteristics of the parts and the technical requirements inferred from the equipment operating conditions, the fields of the inquiry form are reconstructed, price-sensitive information is removed, and a targeted inquiry form for suppliers is generated;

[0033] The category characteristics inherent to a part refer to the default technical requirements inherent to that part category. These are determined by the part category itself and are unrelated to the specific remarks of a particular part. For example, the category characteristics of a head include: normalized state, UT inspection, and minimum thickness guarantee after molding. As long as a part is classified into a certain category, it will automatically have these default requirements.

[0034] S3.5 Establish a method and mechanism for managing inquiry form versions and price data versions, generate a unique inquiry form number, and record the status flow information of inquiry-quotation-order.

[0035] In a preferred embodiment of the present invention, in step S1:

[0036] Similarity is calculated using the weighted Euclidean distance formula.

[0037]

[0038] in, These are weighting coefficients, obtained through training with historical transaction data, reflecting the sensitivity of each parameter to the impact on price. V is the dimensionless normalization coefficient. query V is the current operating condition feature vector (query vector) of the equipment to be queried. i V represents the operating condition feature vector from the i-th historical record in the historical weight estimation table, where j is the j-th dimension of the operating condition feature vector (e.g., pressure, temperature, medium, etc.). query [j] represents the parameter value of the device currently seeking a quote in the j-th dimension, V i [j] represents the parameter value of the i-th historical record in the j-th dimension;

[0039] Dynamically weighted generation of baseline weight Where BaseWeight is the baseline weight (output) of the equipment currently being inquired about, Weight i To estimate the weight of the i-th historical record in the historical weight estimation table, Similarity i Let be the similarity between the i-th historical record and the current device.

[0040] In a preferred embodiment of the present invention, in step S2, the equipment type-part configuration knowledge graph is generated based on the industry design standard HG / T 20570 and historical project statistics;

[0041] The diameter-connector specification mapping rule includes: DN3600 equipment automatically maps main process connector 42", auxiliary connector 36", instrument connector 6", and bottom connector 3".

[0042] The pressure-flange rating mapping rule includes: design pressure < 2.0 MPa maps to 150# flange, design pressure 2.0-5.0 MPa maps to 300# flange, and the flange material rating is checked when the temperature > 300℃;

[0043] The mapping rules for media properties and gasket types include: oxidizing media are mapped to spiral wound gaskets (ASME 16.20A), hydrogen-containing environments are mapped to octagonal gaskets (API 6A), and food-grade media are mapped to polytetrafluoroethylene gaskets.

[0044] In a preferred embodiment of the present invention, in step S2, the component comprehensive coefficient decomposition model is expressed as:

[0045] Price coefficient (comprehensive coefficient) = (material forming coefficient + processing difficulty coefficient + quality grade coefficient) × batch discount coefficient;

[0046] Each part has at least one set of basic coefficients and coefficient adjustment values ​​corresponding to different processes, so as to obtain the final price coefficient based on the part and its process;

[0047] For example:

[0048] The final coefficient for the DN3600×14 end cap using HIC is 1.40 = basic coefficient 1.25 + process coefficient 0.15; the basic coefficient for the DN3600×14×15000 cylinder is 1.05; the basic coefficient for the 42"×60 nozzle is 1.60. If the nozzle wall thickness is 60mm, then the final coefficient for the nozzle is 2.00 = 1.60 + 0.40.

[0049] When using a thick-walled cylinder, the processing difficulty coefficient increases by 0.15; when adding 100% X-ray inspection, the quality grade coefficient increases by 0.30; when adding HIC testing, the quality grade coefficient increases by 0.40; when the quantity is ≥5 pieces, the batch discount coefficient is 0.95.

[0050] The comprehensive coefficient of the end cap is 1.4 = (the basic coefficient of the end cap is 1.25 (fixed) + the quality grade HIC (+0.15) + the processing difficulty is no special (+0)) * the batch discount 1 (less than 5 pieces).

[0051] In a preferred embodiment of the present invention, in step S3, the multi-keyword parallel recognition algorithm adopts a category code system, and each category contains multiple recognition keywords and synonyms;

[0052] in:

[0053] Keywords for end caps include "end cap, head, elliptical end cap, dished end cap", category code CAT-01;

[0054] Keywords for the cylindrical body category include "cylinder, Shell, shell", and the category code is CAT-02;

[0055] Flange-related keywords include "flange, flange, weld neck flange, WN RF", category code CAT-03;

[0056] Keywords for gaskets include "gasket, gasket, spiral wound gasket, octagonal gasket", category code CAT-04;

[0057] Keywords for the nozzle category include "nozzle, flange, forged nozzle", category code CAT-05;

[0058] The keywords for support structure include "saddle, skirt, and skirt", with the category code CAT-06.

[0059] In a preferred embodiment of the present invention, the parameters obtained by parsing the regular expression in step S3 include one or more of the following: diameter, thickness, length, pressure rating, and standard; wherein:

[0060] Head / Cylinder Specification Pattern: DN (diameter) × (thickness) × (length), Example: DN3600×14×15000 is interpreted as a diameter of 3600mm, a wall thickness of 14mm, and a length of 15000mm;

[0061] Flange / Gasket Specification Pattern: (Diameter) × (Pressure Rating) × (Standard Number), Example: 1050-150# ASME16.20A can be interpreted as diameter 1050, pressure rating 150#, and standard ASME;

[0062] Connector specifications: Φ (diameter) × (thickness) × (length), example: Φ1066×60×350 can be interpreted as a diameter of 1066mm, a wall thickness of 60mm, and a length of 350mm.

[0063] In a preferred embodiment of the present invention, in step S3, the technical requirements inference is based on: keyword recognition of the BOM table remarks field and category default technical requirements; wherein, the remarks content is automatically extracted from the original input (S1.1 basic parameters), but the remarks field can also be supplemented or modified after the BOM is generated;

[0064] Keyword identification for the BOM table remarks field includes: HIC marking triggers HIC test qualification requirements, internal coating marking triggers internal surface coating requirements, and T212 marking triggers requirements according to T212 specifications.

[0065] The default technical requirements for each category include: head type, default normalized condition, UT test; flange / pipe type, default Grade III forging, UT / MT test; gasket type, default with inner and outer rings, 304 stainless steel strip + flexible graphite filler strip.

[0066] In a preferred embodiment of the present invention, in step S3, reconstructing the fields of the inquiry form includes:

[0067] Remove sensitive information from the original BOM table. Sensitive information includes unit weight, coefficient, unit price, and subtotal price.

[0068] The original BOM is the complete quotation details table generated by S2.1 (for internal use, containing price-sensitive information); the Request for Quotation is the external version after desensitization and reconstruction by S3.4 (price information is removed, and only necessary technical and business information is retained).

[0069] Retain and standardize the following fields: Part Name, Specification Description, Material, Quantity, Technical Requirements, and Request for Quotation Number;

[0070] Add supplier-targeting fields: recommended supplier, supplier quote, response deadline, and quote validity period.

[0071] In a preferred embodiment of the present invention, the method and mechanism for managing inquiry form version and price data version includes: inquiry form number generation rules, status transition mechanism, and price database automatic update mechanism;

[0072] in:

[0073] Inquiry number generation rules: INQ-[Equipment tag number]-[Category code]-[Date]-[Version], Example: INQ-R201-CAT04-20260325-V1;

[0074] Status progression: Created → Sent to supplier → Supplier has quoted → Price approved → Adopted / Deprecated;

[0075] Automatic price database update: After the supplier returns a quotation, the unit price is parsed and extracted to update the material price database, triggering a recalculation of quotations for related equipment.

[0076] A pressure vessel intelligent quotation and inquiry management system, comprising:

[0077] The weight estimation module is used to receive basic equipment parameters and generate a weight estimation table based on the operating condition feature vector and historical template matching.

[0078] The quotation details generation module is used to automatically extract and generate a BOM (Bill of Materials) from the weight estimation table based on the equipment type-part configuration knowledge graph and the parameter-part mapping rule base.

[0079] The intelligent classification and extraction module is used to classify BOM parts and extract structured specification parameters based on multi-keyword recognition algorithms and regular expression parsing.

[0080] The supplier matching module is used to match the supplier database based on the part category and determine the default supplier and the dedicated inquiry template;

[0081] The Request for Quotation (RFQ) generation module is used to reconstruct RFQ fields, remove price-sensitive information, infer technical requirements, and generate targeted RFQs.

[0082] The version management module is used to generate inquiry numbers, record status progress, update the price database, and track historical price changes.

[0083] The beneficial effects of this invention are: it realizes full automation from weight estimation to supplier inquiry, shortens the traditional quotation-inquiry process that takes 4-6 hours to minutes, and at the same time ensures data consistency and the security of trade secrets. Attached Figure Description

[0084] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:

[0085] Figure 1 This is the overall system architecture diagram of this application (three-layer architecture: weight estimation layer, quotation details layer, and inquiry form layer);

[0086] Figure 2 This is the intelligent generation flowchart for weight estimation in this application (working condition feature vector → historical matching → dynamic weighting → real-time correction).

[0087] Figure 3 This is a flowchart of the parallel recognition of multiple keywords and parsing of regular expressions in this application;

[0088] Figure 4 This is a flowchart of the status transition of the inquiry form in this application. Detailed Implementation

[0089] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0090] Please see Figure 1-4 The embodiments of the present invention include:

[0091] This invention adopts a three-tier architecture:

[0092] First layer: Intelligent generation layer for weight estimation

[0093] Construct a working condition feature vector V=[P, T, M, V, S, C, G], and achieve second-level weight estimation by matching historical templates based on weighted Euclidean distance.

[0094] Second layer: Automatic breakdown layer of price details

[0095] Based on the equipment type-part configuration knowledge graph and parameter-part mapping rule base, a detailed BOM table is automatically generated and part-level price coefficients are calculated.

[0096] Third layer: Product category inquiry form generation layer

[0097] Based on multi-keyword recognition and regular expression parsing, the system intelligently extracts categorized components, matches them with the supplier database, and reconstructs fields to generate targeted inquiry forms.

[0098] A method for intelligent quotation and inquiry management of pressure vessels, comprising the following steps:

[0099] S1. Intelligent generation of weight estimation table

[0100] Step S1-1: Obtain or extract the basic parameters of the equipment to be quoted. The basic parameters include one or more of the following: equipment tag number, equipment name, design pressure, design temperature, medium characteristics, material grade, nominal diameter, tangent length, or design standard.

[0101] The sources of basic parameters include images in different formats such as email, PDF, and JPG, or documents in formats such as Word and Excel.

[0102] Step S1-2: Construct a working condition feature vector V = [P, T, M, V, S, C, G] for querying based on the basic parameters.

[0103] P: Design pressure (MPa)

[0104] T: Design temperature (°C)

[0105] M: Media type (standardized coding)

[0106] V: Volume / External Dimensions

[0107] S: Design standards (ASME VIII-1 / GB150 / NB / T47003, etc.).

[0108] C: Material category (carbon steel / stainless steel / non-ferrous metals / composite materials).

[0109] G: Special process requirements (lining / heat exchange / agitation / jacket / HIC testing, etc.).

[0110] Step S1-3: Historical Template Matching

[0111] The similarity between the load condition feature vector and each record in the historical weight estimation table is calculated using an improved weighted Euclidean distance:

[0112] ,

[0113] in, The weighting coefficients are obtained through training with historical transaction data and reflect the sensitivity of each parameter to the price (pressure weight > temperature weight > diameter weight). V is the dimensionless normalization coefficient. query V is the current operating condition feature vector (query vector) of the equipment to be queried. i V represents the operating condition feature vector from the i-th historical record in the historical weight estimation table, where j is the j-th dimension of the operating condition feature vector (e.g., pressure, temperature, medium, etc.). query [j] represents the parameter value of the device currently seeking a quote in the j-th dimension, V i [j] represents the parameter value of the i-th historical record in the j-th dimension.

[0114] Based on a preset similarity threshold, historical records that match the working condition feature vector and meet the similarity threshold are selected. If multiple historical records are selected, they can be sorted in descending order of similarity, and then the most similar historical record is selected based on a preset quantity threshold.

[0115] Step S1-4: Dynamically weighted generation of benchmark estimation

[0116] Set a similarity threshold θ (default 0.15), filter out the top-K similar historical records, and dynamically weight them to generate the baseline weight of the device to be inquired about:

[0117]

[0118] Where BaseWeight is the base weight of the equipment currently being inquired about, Weight i To estimate the weight of the i-th historical record in the historical weight estimation table, Similarity i Let be the similarity between the i-th historical record and the current device.

[0119] Step S1-5: Apply a real-time correction factor to correct the baseline weight estimation table to generate the final weight estimation table.

[0120] Real-time correction factors include: material price index (based on Shanghai steel spot market API), exchange rate volatility coefficient (for imported materials), and capacity load coefficient (current production schedule of this plant).

[0121] S2. Automatic breakdown of quotation details

[0122] Step S2-1: Equipment type identification and parts configuration

[0123] As shown in the table below, a knowledge graph of equipment type and part configuration is established:

[0124]

[0125] Based on the equipment type-parts configuration knowledge graph, all parts of the equipment to be quoted are automatically extracted from the final weight estimation table, and a detailed Bill of Materials (BOM) is generated.

[0126] The BOM table includes multiple category fields, and the disassembled parts or parameter information is filled into the corresponding category fields. Category fields include part name, specification, material, quantity, unit weight, coefficient, unit price, subtotal, and remarks.

[0127] Step S2-2: Establish parameter-part mapping rules to automatically determine the material standards and dimensional specifications of each part.

[0128] a) Diameter-connector specification mapping rule (based on HG / T 20570 and historical statistics):

[0129] DN3600 equipment → Main process pipe 42" (DN1050) for material inlet and outlet;

[0130] → Auxiliary pipe 36" (DN900) is used for heat exchange medium;

[0131] → Instrument connector 6" (DN150) for temperature / pressure measurement;

[0132] → Bottom connector 3" (DN80) for drainage / sampling.

[0133] b) Pressure-flange rating mapping rule:

[0134] Design pressure < 2.0 MPa → 150# flange (ASME B16.5 Class 150);

[0135] Design pressure 2.0-5.0MPa → 300# flange;

[0136] Temperature > 300℃ → Flange material grade needs to be checked (LF2 CL1 → CL2).

[0137] c) Medium characteristics - gasket type mapping rules:

[0138] Oxidation reactor medium → Spiral wound gasket ASME 16.20A, with inner and outer rings;

[0139] Hydrogen-containing environment → Octagonal / oval pad API 6A;

[0140] Food-grade → Polytetrafluoroethylene gasket;

[0141] Sulfur-containing medium → Flexible graphite spiral wound gasket, inner ring 304 + outer ring carbon steel.

[0142] Step S2-3: Calculation of component comprehensive coefficient

[0143] 1) Establish a coefficient decomposition model: Comprehensive coefficient = (Material forming base value + Processing difficulty adjustment + Quality grade adjustment) × Batch discount coefficient.

[0144] The definition of the component comprehensive coefficient decomposition model is shown in the table below.

[0145]

[0146] 2) Calculate the price coefficient for each part based on the coefficient decomposition model.

[0147] Price coefficients are used to calculate internal quotation costs (subtotals), which directly affect the calculation of quotation amounts. The calculated subtotal data needs to be recorded in the BOM table. Subtotals refer to quotations provided to customers and are considered sensitive information. They will only be reflected in the complete BOM table and do not need to be sent to suppliers. Therefore, they need to be removed from the inquiry form.

[0148] The formula for calculating the subtotal is: Subtotal = Unit weight × Coefficient × Unit price. The higher the coefficient, the greater the proportion of the processing cost of the part (e.g., 1.40 for the end cap includes mold amortization + 15% forming loss, and 2.00 for the nozzle includes forging procurement + machining cost).

[0149] S3. Intelligent Product Category Extraction and Inquiry Form Generation

[0150] Step S3-1: Parallel recognition of multiple keywords

[0151] As shown in the table below, a component category code system is established:

[0152]

[0153] Based on preset part categories, all parts in the BOM are categorized. Part categories include: heads, cylinders, flanges, gaskets, pipes, and support structures.

[0154] Step S3-2: Regular Expression Specification Parsing

[0155] 1) Create a regular expression for each part category.

[0156] a. Head / Cylinder Specifications:

[0157] Regular expression: DN(\d+)[×Xx\*](\d+)(?:[×Xx\*](\d+))?;

[0158] Example: DN3600×14×15000;

[0159] Matching: Diameter = 3600mm, Wall thickness = 14mm, Length = 15000mm.

[0160] b. Flange / Gasket Specifications:

[0161] Regular expression: (\d+)[-\s]?(\d+#|Class\s*\d+);

[0162] Example: 1050-150# ASME 16.20A;

[0163] Matching: Diameter = 1050, Pressure Rating = 150#, Standard = ASME.

[0164] c. Takeover specifications and patterns:

[0165] Regular expression: [ΦD]N?(\d+)[×Xx\*](\d+)[×Xx\*](\d+);

[0166] Example: Φ1066×60×350;

[0167] Matching: Diameter = 1066mm, Wall thickness = 60mm, Length = 350mm.

[0168] 2) Based on the regular expression above, perform structured parsing on the part specification field of the equipment to be quoted obtained in step S2-2, and extract parameters such as nominal diameter, wall thickness, length, pressure rating, and standard number of the part.

[0169] Step S3-3: Based on the part category, match the supplier database to determine the default supplier and dedicated inquiry template for each part.

[0170] Specifically, it includes:

[0171] (1) Matching suppliers: Based on the part category code (CAT-01~CAT-06), search the supplier database for suppliers that can provide this type of part;

[0172] (2) Determine the default supplier: Each part category has a default supplier (as shown in the table below), and the default supplier is selected first;

[0173] (3) Call the dedicated inquiry template: different categories have corresponding templates (such as the gasket type inquiry form format is different from the flange type), which are automatically applied.

[0174] The supplier database contains pre-set supplier information for each category, including supplier name, supply capacity, historical cooperation records, contact information, etc.

[0175] The dedicated inquiry templates are customized according to the characteristics of the part categories. Different categories of inquiry templates contain different required fields and technical requirement formats.

[0176] The system automatically fills the "Recommended Supplier" field of the inquiry form with the matched default supplier, which the user can manually adjust in subsequent steps.

[0177] Step S3-3: Intelligent Inference of Technical Requirements

[0178] The intelligent inference of technical requirements is based on: keyword recognition in the remarks field of the BOM table and default category technical requirements.

[0179] The BOM table's remarks field mainly includes one or more of the keywords listed in the table below:

[0180]

[0181] The default technical requirements for each category include:

[0182] a. End caps: Normalized state, UT testing, minimum thickness guarantee after molding;

[0183] b. Flanges / connectors: Grade III forgings, UT / MT inspection, machining Ra6.3;

[0184] c. Gaskets: with inner and outer rings, 304 stainless steel strip + flexible graphite filler strip, pre-stress control.

[0185] Therefore, the technical requirements inference is based on:

[0186] 1) Keyword identification for BOM remarks field: HIC mark triggers HIC test qualification requirements, internal coating mark triggers internal surface coating requirements, T212 mark triggers requirements according to T212 specification;

[0187] 2) Default technical requirements for each category: Heads are normally heated and UT tested by default; Flanges / pipes are Grade III forgings and UT / MT tested by default; Gaskets are equipped with inner and outer rings and are made of 304 stainless steel strip with flexible graphite filler strip by default.

[0188] Step S3-4: Reconstructing and de-identifying fields in the inquiry form

[0189] Field reconstruction and desensitization mainly include the following processes:

[0190] a) Remove sensitive information such as unit weight, coefficient, unit price, and subtotal price from the original BOM table;

[0191] b) Retain and standardize the following fields: part name, specifications, material, quantity, technical requirements, inquiry number, etc.

[0192] c) Add supplier-oriented fields: recommended supplier, supplier quotation, response deadline, quotation validity period, etc.

[0193] Step S3-5: Generate a targeted quotation form for suppliers based on the information after field reconstruction and desensitization.

[0194] The original BOM lines are compared with the inquiry line lines in the table below:

[0195]

[0196] Step S3-6: Establish a mechanism for managing inquiry form versions and price data versions, generate a unique inquiry form number, and record the inquiry-quotation-order status flow. The inquiry form number can be directly displayed within the inquiry form.

[0197] a. Request for Quotation Number Generation Rules: INQ-[Equipment Tag Number]-[Category Code]-[Date]-[Version].

[0198] Example: INQ-R201-CAT04-20260325-V1

[0199] b. Figure 4 As shown, the status transition includes: Created → Sent to supplier → Supplier has quoted → Price approved → Adopted / Deprecated.

[0200] c. Automatic price database update mechanism: After the supplier returns the quotation, the unit price is parsed and extracted to update the material price database, triggering a recalculation of the quotation for related equipment.

[0201] For example, the automatic price database update mechanism can be seen as follows:

[0202] Supplier returns quotation → OCR / parse to extract unit price → Update material price database → Trigger recalculation of quotation for associated equipment → Prompt for historical quotation deviation

[0203] Among them, the price coefficient can help quickly determine whether the supplier's quotation is reasonable - if the supplier's quotation deviates too much from the subtotal data of the internal estimate, the system will issue a warning.

[0204] Example 1: Fully Automated Process of Oxidation Reactor R201

[0205] Input: Customer inquiry form (PDF attached to email)

[0206] "Oxidation reactor, tag number R-201, diameter 3.6 meters, tangential length 15 meters, design pressure 1.0 MPa, design temperature 200℃, material SA516 Gr.70, medium oxidation reaction liquid, requires HIC testing, internal coating."

[0207] Phase 1: Weight estimation (3 seconds)

[0208] 1. Construct the query condition feature vector: V=[1.0, 200, Oxidation_Liquid, 50m³, ASME,SA516Gr.70, HIC+Coating];

[0209] 2. Fingerprint matching: Material SA516Gr.70, pressure ~1.0MPa, diameter DN3600 → Match all records in the historical weight estimation table (similarity threshold 94%).

[0210] 3. Dynamic weighting: Three similar historical records were selected, and the weights corresponding to these three historical records were extracted to be 22-26 tons. The weighted average weighting based on similarity was used to calculate the generated baseline weight of 24.2 tons.

[0211] 4. Real-time correction: Material price index +3.2% → Adjusted estimated weight 24.2 tons (estimated material cost).

[0212] The 3.2% figure represents the material price index (based on the Shanghai Steel Spot Market API), indicating a 3.2% increase in material prices. This is real-time market data, automatically obtained by the system from an external API. The current material price is then compared with historical benchmark prices. This value may vary each time a quote is submitted, and manual input is not required.

[0213] The "adjusted estimated weight of 24.2 tons" here is the baseline weight of 24.2 tons generated by the dynamic weighting in S1.4 above. The material price index adjustment does not change this value; it is only used to calculate the estimated material cost.

[0214] Phase 2: BOM Disassembly (5 seconds)

[0215] 1. Equipment type identification: "Oxidation reactor" → Based on the equipment type-part configuration knowledge graph, reactor parts include stirring interface, temperature measuring port, feed port, and discharge port;

[0216] 2. Determine the material standards and dimensional specifications of each part based on the parameter-part mapping rule library.

[0217] a) Deduction of the connection specifications:

[0218] • Main feed: 42" (matching the large flow rate requirement of a 3.6-meter diameter);

[0219] • Heat exchange medium: 36" (jacketed or built-in coil interface);

[0220] • Temperature / pressure measurement: 6" (standard instrument interface);

[0221] • Bottom discharge / sampling: 3"×4 (multiple outlets evenly distributed);

[0222] b) Flange rating: 1.0MPa < 2.0MPa → 150# (ASME B16.5);

[0223] d) Gasket standards: Oxidizing media + HIC requirements → Spiral wound gasket ASME 16.20A, with inner and outer rings;

[0224] d) Material upgrade: cylinder / head SA516Gr.70N (normalized, meets HIC), pipe forging SA350 LF2 CL1.

[0225] The generated BOM consists of 21 items, consistent with the structure of the template "MATERIAL LIST FOR MAIN PARTS".

[0226] Phase 3: Coefficient Calculation (2 seconds)

[0227]

[0228] Phase 4: Categorization, Extraction, and Quotation Generation (3 seconds)

[0229] The system automatically identifies 6 parts categories and generates 6 targeted inquiry forms.

[0230] Example of a CAT-04 gasket inquiry:

[0231]

[0232] Anonymization verification: The inquiry form does not contain fields for unit weight, coefficient, unit price, and subtotal, so the supplier cannot deduce the cost structure.

[0233] Phase 5: Version Management and Tracking

[0234] • Inquiry status: Created → Sent to Supplier D → Supplier has provided a quote (return within 48 hours);

[0235] Supplier D returned a quote: DN1050 spiral wound gasket unit price ¥1150 (down 8.7% from the historical price of ¥1260);

[0236] • Price database update: Triggered recalculation of the price for associated equipment R201, material cost reduced by ¥110;

[0237] • Status transition: Price approved → Adopted → Purchase order PO-R201-CAT04-001 generated.

[0238] Total time: 13 seconds (traditional manual methods take 4-6 hours)

[0239] Example 2: Bulk Price Inquiry for Multiple Devices

[0240] A certain project includes 10 pressure vessels (3 storage tanks, 4 reactors, 2 heat exchangers, and 1 tower). Traditionally, it would take 2-3 days to process the inquiry form.

[0241] Invention solution:

[0242] 1. Batch import basic parameters of 10 devices (Excel template);

[0243] 2. Generate 10 weight estimation tables in parallel (total time < 30 seconds);

[0244] 3. Automatically generate 10 BOM tables (total time < 50 seconds);

[0245] 4. Intelligent extraction and aggregation: Identify all 47 CAT-04 gaskets, classify them into 8 types according to specifications, and merge them to generate 8 gasket-related inquiry forms (avoiding duplicate inquiries from the same supplier).

[0246] Generate a total of 6 categories × 10 devices = 32 optimized inquiry forms (the traditional method requires 60-80 forms).

[0247] Supplier collaboration: The same supplier receives combined inquiries, submits batch quotes, and further reduces the unit price (batch discount factor takes effect).

[0248] Example 3: Price Data Version Management Method and Dynamic Adjustment

[0249] For heat exchanger E-101 in historical project P2025-008, the price of 304 stainless steel plate was ¥18.5 / kg when inquiring in June 2025. When inquiring about similar equipment in March 2026, the price rose to ¥22.3 / kg (+20.5%).

[0250] The system will trigger automatically:

[0251] 1. Price database alert: If the price fluctuation of 304 stainless steel exceeds 20%, the historical template price will be marked as invalid.

[0252] 2. Similarity matching adjustment: Reduce the weight of historical templates and increase the weight of real-time price index.

[0253] Price Revision: The base price for the new equipment E-102 will be automatically increased by 18%, and a note will be added to the price list stating "Material prices are high, it is recommended to lock in the purchase or use alternative material 316L".

[0254] The beneficial effects of the intelligent quotation and inquiry management system and method for pressure vessels of this invention are: it can automatically generate a weight estimation table based on the overall parameters of the equipment, generate a detailed quotation BOM table by disassembly, and intelligently extract and classify to generate supplier inquiry forms. Specifically:

[0255] 1. Efficiency Improvement

[0256]

[0257] 2. Improved accuracy: Historical template reuse rate increased from 20% to 85%, inquiry form entry error rate decreased from 8-15% to <1%, and BOM-duplicate table consistency verification coverage reached 100%;

[0258] 3. Data anonymization: Price-sensitive information (unit weight, coefficient, unit price, subtotal) is automatically removed from the inquiry form. Suppliers only receive necessary technical information and cannot reverse-engineer the cost structure. Version management ensures that price changes are traceable.

[0259] 4. Supply chain data collaborative processing: Automatic supplier matching reduces manual allocation errors, dedicated inquiry templates ensure complete transmission of technical requirements, and the price database is automatically updated to guarantee competitive pricing.

[0260] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for intelligent offer and inquiry management of pressure vessels, characterized in that, Includes the following steps: Intelligent generation of S1 weight estimation table S1.1 Obtain or extract the basic parameters of the equipment to be quoted. The basic parameters include one or more of the following: equipment tag number, equipment name, design pressure, design temperature, medium characteristics, material grade, nominal diameter, tangential length, or design standard. S1.2 Construct the current operating condition feature vector V=[P, T,M, V, S, C, G] of the equipment to be queried based on the acquired or extracted basic parameters for easy querying. Where P is the design pressure, T is the design temperature, M is the medium type, V is the volume, S is the design standard, C is the material type, and G is the special process requirements. S1.3 Calculate the similarity between the current working condition feature vector and each record in the historical weight estimation table based on the weighted Euclidean distance, and filter out the historical records that match the working condition feature vector according to the preset similarity threshold, and dynamically generate the benchmark weight estimation table of the equipment to be inquired about. S1.4 Adjust the baseline weight estimation table according to the real-time correction factors to generate the final weight estimation table. The real-time correction factors include the material price index, exchange rate fluctuation coefficient, and capacity load coefficient. Automatic decomposition of S2 quotation details S2.1 Based on the equipment type-part configuration knowledge graph, automatically extract and generate a detailed Bill of Materials (BOM) from the final weight estimation table; S2.2 Based on the parameter-part mapping rule base, automatically determine the material standards and size specifications of each part in the BOM table; S2.3 Calculate the price coefficient of each part in the BOM table based on the comprehensive coefficient decomposition model of the parts, calculate the subtotal data of the equipment to be quoted through the price coefficient, and store the subtotal data in the BOM table; wherein, the decomposition model includes: material forming coefficient, processing difficulty coefficient, quality grade coefficient, and batch discount coefficient; S3 Product Category Intelligent Extraction and Inquiry Form Generation Steps S3.1 Based on a multi-keyword parallel recognition algorithm, the parts in the BOM are classified into preset part categories; S3.2 Use regular expressions to perform structured parsing of the content in the part specification field of the BOM table and extract part parameters; S3.3 Based on the part category, match the supplier database to determine the default supplier and the dedicated inquiry template; S3.4 Based on the inherent category characteristics of the parts and the technical requirements inferred from the equipment operating conditions, the fields of the inquiry form are reconstructed, price-sensitive information is removed, and a targeted inquiry form for suppliers is generated; S3.5 Establish a method and mechanism for managing inquiry form versions and price data versions, generate a unique inquiry form number, and record the status flow information of inquiry-quotation-order.

2. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, In step S1: Similarity is calculated using the weighted Euclidean distance formula. ， in, These are weighting coefficients, obtained through training with historical transaction data, reflecting the sensitivity of each parameter to the impact on price. V is the dimensionless normalization coefficient. query V is the current operating condition feature vector (query vector) of the equipment to be queried. i V represents the operating condition feature vector from the i-th historical record in the historical weight estimation table, where j is the j-th dimension of the operating condition feature vector (e.g., pressure, temperature, medium, etc.). query [j] represents the parameter value of the device currently seeking a quote in the j-th dimension, V i [j] represents the parameter value of the i-th historical record in the j-th dimension; Dynamically weighted generation of baseline weight Where BaseWeight is the base weight of the equipment currently being inquired about, and Weight i To estimate the weight of the i-th historical record in the historical weight estimation table, Similarity i Let be the similarity between the i-th historical record and the current device.

3. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, In step S2, the equipment type-part configuration knowledge graph is generated based on the industry design standard HG / T 20570 and historical project statistics; The diameter-connector specification mapping rule includes: DN3600 equipment automatically maps main process connector 42", auxiliary connector 36", instrument connector 6", and bottom connector 3". The pressure-flange rating mapping rule includes: design pressure < 2.0 MPa maps to 150# flange, design pressure 2.0-5.0 MPa maps to 300# flange, and the flange material rating is checked when the temperature > 300℃; The mapping rules for media properties and gasket types include: oxidizing media are mapped to spiral wound gaskets (ASME 16.20A), hydrogen-containing environments are mapped to octagonal gaskets (API 6A), and food-grade media are mapped to polytetrafluoroethylene gaskets.

4. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, In step S2, the component comprehensive coefficient decomposition model is expressed as: Price coefficient (comprehensive coefficient) = (material forming coefficient + processing difficulty coefficient + quality grade coefficient) × batch discount coefficient; Each part has at least one set of basic coefficients and coefficient adjustment values ​​corresponding to different processes, so as to obtain the final price coefficient based on the part and its process.

5. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, In step S3, the multi-keyword parallel recognition algorithm adopts a category code system, with each category containing multiple recognition keywords and synonyms; in: Keywords for end caps include "end cap, head, elliptical end cap, dished end cap", category code CAT-01; Keywords for the cylindrical body category include "cylinder, Shell, shell", and the category code is CAT-02; Flange-related keywords include "flange, flange, weld neck flange, WN RF", category code CAT-03; Keywords for gaskets include "gasket, gasket, spiral wound gasket, octagonal gasket", category code CAT-04; Keywords for the nozzle category include "nozzle, flange, forged nozzle", category code CAT-05; The keywords for support structure include "saddle, skirt, and skirt", with the category code CAT-06.

6. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, The parameters obtained from the regular expression specification parsing in step S3 include one or more of the following: diameter, thickness, length, pressure rating, and standard; wherein: Head / Cylinder Specification Pattern: DN Diameter × Thickness × Length; Flange / Gasket Specification Pattern: Diameter × Pressure Rating × Standard Number; Connector specifications: Φ diameter × thickness × length.

7. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, In step S3, the technical requirements are inferred based on: keyword identification in the remarks field of the BOM table and the default category technical requirements; Keyword identification for the BOM remarks field includes: HIC mark triggers HIC test qualification requirements, internal coating mark triggers internal surface coating requirements, and T212 mark triggers requirements according to T212 specification. The default technical requirements for each category include: head type, default normalized condition, UT test; flange / pipe type, default Grade III forging, UT / MT test; gasket type, default with inner and outer rings, 304 stainless steel strip + flexible graphite filler strip.

8. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, In step S3, the reconstruction of the inquiry form fields includes: Remove sensitive information from the original BOM table. Sensitive information includes unit weight, coefficient, unit price, and subtotal price. Retain and standardize the following fields: Part Name, Specification Description, Material, Quantity, Technical Requirements, and Request for Quotation Number; Add supplier-targeting fields: recommended supplier, supplier quote, response deadline, and quote validity period.

9. The intelligent quotation and inquiry management method for pressure vessels according to claim 1, characterized in that, The methods and mechanisms for managing inquiry form versions and price data versions include: inquiry form number generation rules, status transition mechanism, and automatic price database update mechanism. in: Inquiry number generation rules: INQ-[Equipment tag number]-[Category code]-[Date]-[Version]; Status progression: Created → Sent to supplier → Supplier has quoted → Price approved → Adopted / Deprecated; Automatic price database update: After the supplier returns a quotation, the unit price is parsed and extracted to update the material price database, triggering a recalculation of quotations for related equipment.

10. A pressure vessel intelligent quotation and inquiry management system, characterized in that, include: The weight estimation module is used to receive basic equipment parameters and generate a weight estimation table based on the operating condition feature vector and historical template matching. The quotation details generation module is used to automatically extract and generate a BOM (Bill of Materials) from the weight estimation table based on the equipment type-part configuration knowledge graph and the parameter-part mapping rule base. The intelligent classification and extraction module is used to classify BOM parts and extract structured specification parameters based on multi-keyword recognition algorithms and regular expression parsing. The supplier matching module is used to match the supplier database based on the part category and determine the default supplier and the dedicated inquiry template; The Request for Quotation (RFQ) generation module is used to reconstruct RFQ fields, remove price-sensitive information, infer technical requirements, and generate targeted RFQs. The version management module is used to generate inquiry numbers, record status progress, update the price database, and track historical price changes.

Images