# US 12,223,233 B2 · Method and system for computer-aided design of a technical system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,223,233 B2 - Document ID: US12223233B2 - Status: Granted - Kind: B2 - Application number: 16/976,337 - Filed: 2019-02-14 - Published: 2021-05-13 - Granted: 2025-02-11 - Earliest claimed: 2018-02-28 - Assignees: Siemens Aktiengesellschaft - Applicants: Siemens Aktiengesellschaft - Inventors: Jan Fischer; Vincent Malik; Jan Christoph Wehrstedt; Nils Weinert - Web page: https://patentgrep.com/patent/US12223233B2 ## CPC - G06F 30/00: Computer-aided design [CAD] - G06F 30/13: Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads - G06F 16/953: Querying, e.g. by the use of web search engines - G06F 30/10: Geometric CAD - G06F 30/20: Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30) - G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model - G06N 3/08: Learning methods - G06N 3/09: Supervised learning - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A component of the technical system, a component designation and a characteristic parameter designation for a characteristic parameter of relevance to the design of the components are read and a search engine is queried with same is provided. The documents found by the search engine are read and component information, e.g. product information concerning a specific component, is extracted from said documents. The extracted component information is supplied to a machine learning routine which has been trained, using a plurality of predefined training component information and training characteristic parameter values, to reproduce predefined training characteristic parameter values using predefined training component information. Output data from the machine learning routine is selected as characteristic parameter values and inserted into a planning data record. The planning data record is then output to design the technical system. ## Claims ### Claim 1 A method for computer-aided design of a technical system, said method comprising: for a respective component of the technical system; actuating a search engine, said actuating using a search term comprising a component name and/or a characteristic parameter name for a design-relevant characteristic parameter of the component; and extracting component specifications from documents found by the search engine; training a machine learning routine on a basis of a multiplicity of predefined training component specifications and predefined training characteristic parameter values, to reproduce predefined training characteristic parameter values on a basis of predefined training component specifications; executing the trained machine learning routine, said executing using input data and generating output data, wherein the input data comprises the extracted component specifications; generating a planning data record for the technical system, wherein characteristic parameter values are selected from the output data generated from said executing the trained machine learning routine and are inserted into the planning data record during said generating the planning data record, wherein the characteristic parameter values are design relevant characteristic parameters of components; and outputting the planning data record in order to design the technical system. ### Claim 2 (depends on 1) The method as claimed in claim 1, wherein the machine learning routine is implemented by way of an artificial neural network, a recurrent neural network, a convolutional neural network, an autoencoder, a deep learning architecture, a support vector machine, a data-driven trainable regression model, a k-nearest-neighbor classifier, or a decision tree. ### Claim 3 (depends on 1) The method as claimed in claim 1, wherein the selected characteristic parameter values were selected depending on a target value for a design parameter predefined for the technical system. ### Claim 4 (depends on 1) The method as claimed claim 1, wherein the method comprises: configuring a simulator of the technical system by way of the generated planning data record; simulating a technical function of the technical system by way of the configured simulator (SIM); and outputting a functional specification about the simulated technical function. ### Claim 5 (depends on 4) The method as claimed in claim 4, wherein the characteristic parameter values to be inserted into the planning data record are selected such that a deviation of the functional specification from a functional requirement predefined for the technical system is reduced. ### Claim 6 (depends on 1) The method as claimed in claim 1, wherein the method comprises: generating a multiplicity of planning data records each containing characteristic parameter values selected differently from the output data generated from said executing the machine learning routine; simulating a technical function of the technical system by way of a simulator for a respectively generated planning data record; and selecting a planning data record that optimizes the simulated technical function from the multiplicity of generated planning data records in order to design the technical system. ### Claim 7 (depends on 1) The method as claimed in claim 1, wherein the trained machine learning routine was trained in a component type-specific manner; an wherein the method comprises: feeding the extracted component specifications, together with a respective component type specification, to the machine learning routine. ### Claim 8 (depends on 1) The method as claimed in claim 1, wherein: the machine learning routine has been trained, on the basis of a multiplicity of predefined training component specifications, training characteristic parameter names and training characteristic parameter values, to reproduce predefined training characteristic parameter values on the basis of predefined training component specifications and training characteristic parameter names; and the machine learning routine that was trained on the basis of the multiplicity of predefined training component specifications is fed the characteristic parameter names in association with the extracted component specifications. ### Claim 9 (depends on 1) The method as claimed in claim 1, wherein the search engine uses a product data-focused web crawler to search through a data network. ### Claim 10 (depends on 1) The method as claimed in claim 1, wherein the component specifications are extracted from the found documents by way of a parser and/or a pattern recognizer. ### Claim 11 (depends on 1) The method as claimed in claim 1, wherein a respective component name and/or characteristic parameter name is read in from a blueprint of the technical system. ### Claim 12 (depends on 1) The method as claimed in claim 1, further comprising: parameterizing a multiplicity of planning data records that include the generated planning data record for the technical system; and extracting a multiplicity of training characteristic parameter values from the multiplicity of parameterized planning data records. ### Claim 13 (depends on 1) A computer system for computer-aided design of a technical system, said computer system comprising one or more processers configured to perform the method as claimed in claim 1. ### Claim 14 (depends on 1) A computer program product comprising a computer readable hardware storage device storing computer readable program code that is executable by a processor of a computer system to implement the method according to claim 1. ### Claim 15 (depends on 14) The computer program product as claimed in claim 14, wherein a non-transitory computer-readable storage medium comprises the computer program product.