# US 11,897,066 B2 · Simulation apparatus > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,897,066 B2 - Document ID: US11897066B2 - Status: Granted - Kind: B2 - Application number: 16/409,873 - Filed: 2019-05-13 - Published: 2019-11-21 - Granted: 2024-02-13 - Earliest claimed: 2018-05-17 - Assignees: Fanuc Corporation - Applicants: Fanuc Corporation - Inventors: Satoshi Uchida - Web page: https://patentgrep.com/patent/US11897066B2 ## CPC - B23Q 15/013: Control or regulation of feed movement (B23Q15/12 takes precedence) - G05B 13/0265: {the criterion being a learning criterion} - G05B 19/4069: Simulating machining process on screen (G05B19/4068 takes precedence) - G05B 19/41885: {characterised by modeling, simulation of the manufacturing system} - G06N 3/08: Learning methods - G06N 3/092: Reinforcement learning - G05B 2219/33301: Simulation during machining - G05B 2219/39298: Trajectory learning - G05B 2219/42342: Path, trajectory tracking control - Y02P 90/02: Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS] - B23Q: DETAILS, COMPONENTS, OR ACCESSORIES FOR MACHINE TOOLS, e.g. ARRANGEMENTS FOR COPYING OR CONTROLLING (tools of the kind used in lathes or boring machines B23B27/00); MACHINE TOOLS IN GENERAL CHARACTERISED BY THE CONSTRUCTION OF PARTICULAR DETAILS OR COMPONENTS; COMBINATIONS OR ASSOCIATIONS OF METAL-WORKING MACHINES, NOT DIRECTED TO A PARTICULAR RESULT - G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - Y02P: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS ## Abstract A simulation apparatus includes a machine learning device for learning a change in a machining route in machining of a workpiece. The machine learning device observes data indicating the changed machining route and data indicating a machining condition of the workpiece as a state variable, and also acquires determination data for determining whether or not a cycle time obtained by simulation using the changed machining route is appropriate, and learns by associating the machining condition of the workpiece with the change in the machining route, using the state variable and the determination data. ## Claims ### Claim 1 A simulation apparatus for changing a machining route in machining of a workpiece performed in a machine tool to shorten a cycle time for machining the workpiece, the simulation apparatus comprising: a simulation unit, implemented by a processor, for simulating the machining of the workpiece in the machine tool; and a machine learning device for learning a change in the machining route, wherein the machine learning device includes a neural network, and observes (i) after-change machining route data indicating a changed machining route changed based on a command of the machine tool and (ii) machining condition data indicating a machining condition of the workpiece, as a state variable representing a current state of an environment, acquires cycle time determination data to determine whether or not the cycle time for machining the workpiece is appropriate, with a smaller cycle time being appropriate and a larger cycle time being not appropriate, among a result of simulation performed by the simulation unit based on the changed machining route, as determination data indicating propriety determination result of a change in the machining route, performs a multilayer-structure calculation corresponding to the neural network using the state variable and the determination data, and learns a correlation between the machining condition of the workpiece and the change in the machining route, and wherein the change in the machining route includes (a) a change in a direction and a length or (b) a change in a coordinate value of a command unit implemented by the processor and configuring the machining route, and the simulation unit is configured to change the machining of the workpiece performed in the machine tool based on the change learned by the machine learning device, and output a machining command corresponding to the changed machining route to the machine tool to perform the machining of the workpiece. ### Claim 2 (depends on 1) The simulation apparatus according to claim 1, wherein the machine learning device further acquires shock determination data for determining a shock occurring in the machine tool due to machining, among the result of the simulation, as the determination data indicating the propriety determination result of the change in the machining route. ### Claim 3 (depends on 1) The simulation apparatus according to claim 1, wherein the machine learning device obtains a reward related to the propriety determination result, wherein the reward is higher as the cycle time is shorter, and updates a function representing an action to change the machining route with respect to the machining condition of the workpiece, using the reward. ### Claim 4 A simulation apparatus for changing a machining route in machining of a workpiece performed in a machine tool to shorten a cycle time for machining the workpiece, the simulation apparatus comprising: a simulation unit, implemented by a processor, for simulating the machining of the workpiece in the machine tool; and a machine learning device for determining a change in the machining route, wherein the machine learning device includes a neural network, and observes (i) after-change machining route data indicating a changed machining route changed based on a command of the machine tool and (ii) machining condition data indicating a machining condition of the workpiece, as a state variable representing a current state of an environment, performs a multilayer-structure calculation corresponding to the neural network using the state variable, learns a correlation between the machining condition of the workpiece and the change in the machining route, and determines, by using the state variable and the correlation learned by the machine learning device, the change in the machining route, and wherein the change in the machining route includes (a) a change in a direction and a length or (b) a change in a coordinate value of a command unit implemented by the processor and configuring the machining route, and the simulation unit is configured to change the machining of the workpiece performed in the machine tool based on the change learned by the machine learning device, and output a machining command corresponding to the changed machining route to the machine tool to perform the machining of the workpiece.