Systems and methods for instantaneous performance management of machine tools

Abstract

A system (100), apparatus (110), and method for instantaneous performance management of a machine tool (105) are disclosed. The method includes receiving real-time condition data associated with one or more components (225) of the machine tool (105) from one or more sources (115). Furthermore, based on the condition data, at least one parameter value associated with one or more key components that may affect the performance of the machine tool (105) is calculated. Furthermore, based on the parameter values, a digital twin of the machine tool (105) is configured to simulate the behavior of one or more key components in a simulation environment. Furthermore, based on the simulated behavior of one or more key components, the impact on the performance of the machine tool (105) is predicted. Furthermore, based on the predicted impact, the operation of the machine tool (105) is optimized.

Description

[0001] This invention relates to the field of machine tools, and more particularly to systems, apparatus, and methods for real-time performance management of machine tools.

[0002] In many industries, machine tools play a crucial role in shaping rigid materials into suitable forms. Workpiece machining can be performed through various types of operations, such as cutting, milling, shearing, and drilling. Each of these operations is performed by a suitable cutting tool attached to the spindle of the machine tool. The movement of the workpiece relative to the cutting tool is controlled by a guide mechanism. However, vibrations in the components of the cutting tool, spindle, or guide mechanism can cause chatter marks to form on the workpiece. Therefore, the performance of a machine tool is determined by the dynamics of its various components. These dynamics also vary based on the condition of the components. For example, wear on the cutting tool can lead to poor machining quality.

[0003] Given the above, there is a need for instantaneous performance management of machine tools based on real-time status data.

[0004] The works "Digital twin for CNC machine tool: modeling and using strategy" and "Digital twin modeling method for CNC machine tool" by LUO WEICHAO et al. both involve the principle of using digital twins of CNC machine tools.

[0005] Therefore, the object of the present invention is to provide a system, apparatus, and method for instantaneous performance management of machine tools. This object is achieved by the method for instantaneous performance management of machine tools disclosed herein. As used herein, the term "machine tool" refers to a computerized numerical control (CNC) machine. CNC machines can be of different types, such as CNC milling machines, CNC drilling machines, CNC lathes, CNC turning centers, CNC plasma cutting machines, CNC special-purpose machines, and CNC grinding machines. A CNC machine can include any number of axes. For example, the number of axes can be one of 2, 2.5, 3, 4, 5, 9, 10, etc. As used herein, the term "instantaneous performance" refers to the real-time performance of the machine tool. As used herein, the term "performance" refers to a measure that affects the productivity of the machine tool. In this disclosure, performance can be measured based on, for example, machining cycle time, stability, or accuracy.

[0006] This method includes receiving real-time status data associated with one or more components of a machine tool from one or more sources. The one or more sources may include, but are not limited to, the machine tool's controller, sensing units, and edge devices. Sensing units include, but are not limited to, position sensors, rotary encoders, force gauges, proximity sensors, current sensors, accelerometers, temperature sensors, and acoustic sensors. The real-time status data indicates one or more operating conditions of the machine tool in real time. The status data may be associated with sensor data, the machine tool's operating conditions, and specifications. As used herein, the term "sensor data" refers to the output of one or more sensing units associated with the machine tool. The term "operating conditions" includes parameters set by the operator. Operating conditions may include, for example, the type of machining operation, the type of cutting tool, tool settings, automatic tool changer (ATC) settings, feed rate, cutting speed, spindle speed, spindle power, axis torque, axis speed, axis power, strokes per minute for each axis, and stroke range. Furthermore, operating conditions may also include data associated with G-codes, M-codes, and computer-aided design models, based on which the machine tool processes workpieces. As used herein, the term "specification" refers to the specifications associated with the type and maximum capacity of the machine tool provided by the original equipment manufacturer (OEM). Specifications may include, but are not limited to, spindle torque and power, spindle size, number of axes, axis power, stroke, machine tool dimensions, and motor ratings. Similarly, specifications also include the capacities associated with the machine tool's components as specified by the OEM.

[0007] Operating conditions can be set by the machine tool operator or preset by the machine tool's OEM. In addition, condition data may include non-operating parameters, such as material test data associated with the workpiece attached to the machine tool. One or more components include, but are not limited to, bed, column, spindle, cutting tool, spindle motor, ball screw, lead screw, linear motion guide, axis motor, and worktable.

[0008] The method further includes calculating at least one parameter value associated with one or more critical components that may affect machine tool performance based on condition data. In one embodiment, calculating the at least one parameter value associated with one or more critical components includes calculating a stiffness value associated with one or more critical components based on condition data. The term "stiffness value" as used herein refers to the degree to which a component resists deformation. For example, a stiffness value may be expressed as a torsional constant or using Young's modulus. A stiffness value may correspond to at least one of static stiffness and dynamic stiffness. Dynamic stiffness is the ratio between dynamic forces on a component and the component's final dynamic displacement. Static stiffness is the ratio between static forces on a component and the component's final static deflection. Stiffness values ​​provide information about how the stiffness of one or more critical components changes with temperature. For example, during cutting, the spindle stiffness changes with temperature. Changes in stiffness can affect machining accuracy and also affect machine tool performance.

[0009] Advantageously, the present invention uses real-time condition data of the machine tool for real-time calculation of stiffness values ​​of one or more key components.

[0010] The method further includes configuring a digital twin of the machine tool based on parameter values. In one implementation, the digital twin is a dynamic virtual copy based on one or more of the following: a physics-based model, a computer-aided design (CAD) model, a computer-aided engineering (CAE) model, a one-dimensional (1D) model, a two-dimensional (2D) model, a three-dimensional (3D) model, a finite element (FE) model, a descriptive model, a meta-model, a stochastic model, a parametric model, a reduced-order model, a statistical model, a heuristic model, a predictive model, an aging model, a machine learning model, an artificial intelligence model, a deep learning model, a system model, a knowledge graph, etc. In one embodiment, configuring the digital twin of the machine tool includes updating the digital twin of the machine tool based on condition data and calculated parameter values. Updating the digital twin of the machine tool is to replicate a response that is substantially similar to the real-time response of the machine tool through simulation. In other words, the digital twin is configured to represent the real-time state of the machine tool.

[0011] Advantageously, the digital twin acts as a soft sensor, which facilitates the replication of the responses of internal components of a machine tool that cannot be measured using physical sensors.

[0012] The method further includes simulating the behavior of one or more key components in a simulation environment based on a configured digital twin. According to embodiments of the invention, simulating the behavior of one or more key components includes generating simulation instances based on the configured digital twin. The simulation instance may be a simulation thread associated with a simulation model, which operates independently of all other threads during execution. Furthermore, the simulation instance is executed in a simulation environment using the simulation model to generate simulation results indicative of the behavior of one or more key components. The simulation model may be an analysis model in a machine-executable form, derived from at least one of a physics-based model, a data-driven model, or a hybrid model associated with the machine tool. The simulation model may be a one-dimensional (1D) model, a two-dimensional (2D) model, a three-dimensional (3D) model, or a combination thereof. The simulation instance may be executed in the simulation environment as one of stochastic simulation, deterministic simulation, dynamic simulation, continuous simulation, discrete simulation, local simulation, distributed simulation, cooperative simulation, or a combination thereof. It must be understood that the simulation model involved in this disclosure may include both system-level models and component-level models associated with the machine tool. In one embodiment, the behavior of one or more key components is associated with the stability of the machine tool. Stability can be associated with machine tool chatter, cutting tools, shafts or spindles, or the overall structural stability of the machine tool. In another embodiment, the behavior of one or more critical components is associated with one or more defects in one or more critical components. Defects in components are the result of one or more failure modes. Defects include, but are not limited to, lank wear, notch wear, crater wear, plastic deformation, thermal cracking, splintering, and fatigue fracture.

[0013] Advantageously, this invention facilitates the determination of different types of stability and machine tool-related defects through simulation based on digital twins. Because the digital twin replicates the real-time behavior of the machine tool, the simulation results are more accurate compared to traditional analysis techniques based solely on sensor data.

[0014] The method further includes predicting the impact on machine tool performance based on the behavior of one or more key components in a simulation environment. In one embodiment, predicting the impact on machine tool performance includes determining the cycle time associated with the machine tool based on the machine tool's stability. As used herein, the term "cycle time" refers to the time it takes for the machine tool to complete a production run. In other words, the term "cycle time" can be defined as the time it takes for the machine tool to complete a cutting operation. In another embodiment, predicting the impact on machine tool performance includes predicting the impact on machining accuracy associated with the machine tool based on the machine tool's stability. Machining accuracy indicates the difference between actual measurements specified by the operator for the workpiece and measurements of the finished workpiece. Actual measurements can be specified in a CAD file provided to the machine tool. In yet another embodiment, predicting the impact on machine tool performance includes calculating the remaining service life of one or more key components based on one or more defects.

[0015] Advantageously, the present invention facilitates the prediction of the impact on machine tool performance based on real-time condition data.

[0016] The method further includes optimizing machine tool operation based on its impact on machine tool performance. In one embodiment, optimizing machine tool operation includes identifying at least one control parameter for improving machine tool performance based on this impact. Control parameters include, but are not limited to, cutting speed, feed rate, and depth of cut. Furthermore, values ​​of the control parameters are calculated based on their impact on machine tool performance. Additionally, machine tool operation is simulated based on the calculated values ​​of the control parameters. Furthermore, the simulated performance of the machine tool from the simulated operation is compared to a threshold. If the simulated performance of the machine tool is greater than the threshold, the calculated values ​​of the control parameters are applied to operate the machine tool.

[0017] Advantageously, the present invention facilitates the optimization of machine tool operation during runtime based on real-time condition parameters for performance improvement.

[0018] According to embodiments of the invention, the method further includes generating one or more recommendations for improving the design of one or more key components based on their impact on the performance of the machine tool. In one example, the design can be improved by modifying the stiffness associated with the materials of one or more key components of the machine tool. The one or more key components may include, for example, a bed, column, shaft, or spindle associated with the machine tool. In one embodiment, the recommendations may be generated based on a knowledge graph.

[0019] According to embodiments of the invention, the method further includes scheduling maintenance activities for the machine tool based on its impact on performance. These maintenance activities may be associated with preventative maintenance, reactive maintenance, or predictive maintenance.

[0020] The object of this invention is achieved by an apparatus for real-time performance management of a machine tool. The apparatus includes one or more processing units and a memory unit communicatively coupled to the one or more processing units. The memory unit includes a condition management module stored in the form of machine-readable instructions executable by the one or more processing units. The condition management module is configured to perform the method steps described above. Execution of the condition management module can also be implemented using a coprocessor, such as a graphics processing unit (GPU), a field-programmable gate array (FPGA), or a neural processing / computing engine. Furthermore, the memory unit may also include a database.

[0021] According to embodiments of the present invention, the device can be an edge computing device. As used herein, "edge computing" refers to a computing environment capable of operating on an edge device (e.g., connected at one end to a sensing unit in an industrial setup and at the other end to one or more remote servers, such as computing servers or cloud computing servers), which can be a compact computing device with a small form factor and resource constraints in terms of computing power. A network of the edge computing device can also be used to implement the device. Such a network for the edge computing device is called a fog network.

[0022] In another embodiment, the device is a cloud computing system with a cloud-based platform configured to provide cloud services for analyzing machine tool condition data. As used herein, "cloud computing" refers to a processing environment that includes configurable physical and logical computing resources (e.g., networks, servers, storage devices, applications, services, etc.) and data distributed across a network (e.g., the Internet). The cloud computing platform can be implemented as a service for analyzing condition data. In other words, the cloud computing system provides on-demand network access to a shared pool of configurable physical and logical computing resources. This network may be, for example, a wired network, a wireless network, a communication network, or a network formed by any combination of these networks.

[0023] Additionally, the object of the present invention is achieved by a system for instantaneous performance management of a machine tool. This system includes one or more sources capable of providing real-time status data associated with the machine tool, and means configured for instantaneous performance management of the machine tool as described above, the means being communicatively coupled to the one or more sources. As used herein, the term "source" refers to an electronic device configured to acquire status data and transmit the status data to the means. Non-limiting examples of sources include sensing units, controllers, and edge devices.

[0024] The object of the present invention is also achieved by a computer program product. This computer program product may be, for example, a computer program or include another element besides a computer program. This other element may be hardware, such as a memory device on which the computer program is stored, a hardware key for using the computer program, etc., and / or may be software, such as documentation or a software key for using the computer program.

[0025] The above-described properties, features, and advantages of the invention, as well as the ways in which they are realized, will become more apparent and understandable from the following description of embodiments of the invention in conjunction with the accompanying drawings. The described embodiments are intended to illustrate, not limit, the invention.

[0026] The invention will be further described below with reference to the embodiments illustrated in the accompanying drawings, in which;

[0027] Figure 1A The illustration shows a block diagram of a system for instantaneous performance management of a machine tool according to an embodiment of the present invention;

[0028] Figure 1B The illustration shows a block diagram of an apparatus for instantaneous performance management of a machine tool according to an embodiment of the present invention;

[0029] Figure 2 The diagram illustrates a block diagram of a test setup for constructing a digital twin of a machine tool according to an embodiment of the present invention;

[0030] Figure 3 A flowchart illustrating an exemplary method for instantaneous performance management of a machine tool according to embodiments of the present invention is described; and

[0031] Figure 4 A flowchart illustrating an exemplary method for optimizing machine tool operation according to an embodiment of the present invention is provided.

[0032] Embodiments for carrying out the invention are described in detail below. Various embodiments are described with reference to the accompanying drawings, wherein similar reference numerals are used throughout to refer to similar elements. In the following description, numerous specific details are set forth for purposes of explanation in order to provide a thorough understanding of one or more embodiments. It will be apparent that such embodiments can be practiced without these specific details.

[0033] Figure 1AA block diagram of a system 100 for instantaneous performance management of a machine tool 105 according to an embodiment of the present invention is illustrated. System 100 includes a device 110 communicatively coupled via a network 120 to a controller 115 associated with the machine tool 105. Device 110 is an edge computing device. Those skilled in the art will understand that device 110 may be communicatively coupled to multiple controllers in a similar manner. Each controller may be associated with one or more machine tools. Controller 115 enables the operator of the machine tool 105 to define operating conditions for performing machining operations. As those skilled in the art will understand, operating conditions may be defined by the operator before initiating a machining operation or during operation of the machine tool 105. For example, operating conditions correspond to the type of machining operation, cutting tool type, tool settings, automatic tool changer settings, feed rate, cutting speed, G-code, M-code, and material test data associated with a workpiece mounted on the machine tool 105. Controller 115 may further be communicatively coupled to one or more sensing units 125 associated with the machine tool 105. One or more sensing units include at least one of a position sensor, an accelerometer, a rotary encoder, a force gauge, a current sensor, a thermistor, an acoustic sensor, and an image sensor. The position sensor is configured to determine the linear position of a workpiece mounted on machine tool 105. The accelerometer is configured to measure vibrations at one or more locations on machine tool 105. In this embodiment, the accelerometer is mounted on the structure of machine tool 105. For example, the accelerometer may be attached to the bed or column of machine tool 105. The rotary encoder is configured to measure the number of rotations of the spindle on machine tool 105. The force gauge is configured to measure the cutting force associated with a cutting tool on machine tool 105. The current sensor is configured to measure the current associated with a servo mechanism controlling the spindle movement of machine tool 105. The thermistor is configured to measure the temperature at one or more locations on machine tool 105. The acoustic sensor is configured to measure the noise level generated by machine tool 105. The image sensor is configured to capture an image of the cutting tool for determining tool wear.

[0034] The controller 115 includes a transceiver 130, one or more processors 135, and a memory 140. The transceiver 130 is configured to connect the controller 115 to a network interface 145 associated with network 120. The controller 115 transmits real-time status data to the device 110 via the network interface 145. The real-time status data includes operating conditions set on the controller 115 and sensor data received from one or more sensing units 125.

[0035] Device 110 may be a (personal) computer, workstation, virtual machine running on host hardware, microcontroller, or integrated circuit. Alternatively, device 110 may be a real or virtual computer group (the technical term for a real computer group is "cluster," and the technical term for a virtual computer group is "cloud").

[0036] Device 110 includes a communication unit 150, one or more processing units 155, a display 160, a graphical user interface (GUI) 165, and a memory unit 170, which are communicatively coupled to each other. Figure 1B As shown. In one embodiment, communication unit 150 includes a transmitter (not shown), a receiver (not shown), and a gigabit Ethernet port (not shown). Memory unit 170 may include a stacked 2 gigabyte random access memory (RAM) stack-up package (PoP) and flash memory. One or more processing units 155 are configured to execute computer program instructions defined in the module. Furthermore, one or more processing units 155 are also configured to simultaneously execute instructions in memory unit 170. Display 160 includes a high-definition multimedia interface (HDMI) display and a cooling fan (not shown). Additionally, an operator can access device 110 via GUI 165. GUI 165 may include a web-based interface, a web-based downloadable application interface, etc.

[0037] As used herein, processing unit 155 refers to any type of computing circuit, such as, but not limited to, a microprocessor, microcontroller, complex instruction set computing microprocessor, reduced instruction set computing microprocessor, very long instruction word microprocessor, explicit parallel instruction computing microprocessor, graphics processor, digital signal processor, or any other type of processing circuit. Processing unit 155 may also include embedded controllers, such as general-purpose or programmable logic devices or arrays, application-specific integrated circuits, single-chip computers, etc. Typically, processing unit 155 may include hardware and software elements. Processing unit 155 can be configured for multithreading, meaning that processing unit 155 can simultaneously manage different computational processes, thereby executing them in parallel, or switching between active and passive computational processes.

[0038] Memory unit 170 can be volatile or non-volatile memory. Memory unit 170 can be coupled for communication with processing unit 155. Processing unit 155 can execute instructions and / or code stored in memory unit 170. Various computer-readable storage media can be stored in and accessed from memory unit 170. Memory unit 170 can include any suitable elements for storing data and machine-readable instructions, such as read-only memory, random access memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, hard disk drive, removable media drive for handling compact disks, digital video disks, floppy disks, magnetic tape cassettes, memory cards, etc.

[0039] The memory unit 170 further includes a condition management module 175 in the form of machine-readable instructions on any of the aforementioned storage media, and is communicative to and executed by the processing unit 155. The condition management module 175 further includes a preprocessing module 177, a parameter calculation module 180, a digital twin module 182, a simulation module 185, an analysis module 187, an optimization module 190, a recommendation module 192, a maintenance module 195, and a notification module 197. The device 110 may further include a storage unit 198. The storage unit 198 may include a database containing operation or performance logs associated with all machine tools communicatively coupled to the device 110. The following description explains the function of these modules when executed by the processing unit 155.

[0040] The preprocessing module 177 is configured to preprocess real-time status data received from the controller 115. Preprocessing of operational data may include various steps for preparing status data for further processing. These steps may include, but are not limited to, data cleaning, data normalization, and data selection.

[0041] The parameter calculation module 180 is configured to calculate, based on condition data, at least one parameter value associated with one or more key components that may affect the performance of the machine tool 105.

[0042] The digital twin module 182 configures the digital twin of the machine tool 105 based on this parameter value. The digital twin module 182 continuously calibrates the digital twin of the machine tool 105 to replicate a substantially similar real-time response of the machine tool 105 during simulation. In other words, the digital twin is calibrated to ensure a certain degree of fidelity with the machine tool 105.

[0043] The simulation module 185 is configured to simulate the behavior of one or more key components in a simulation environment based on the configured digital twin.

[0044] Analysis module 187 is configured to predict the impact on the performance of machine tool 105 based on the behavior of one or more key components in a simulation environment.

[0045] The optimization module 190 is configured to optimize the operation of the machine tool 105 based on its impact on the performance of the machine tool 105.

[0046] Recommendation module 192 is configured to generate one or more recommendations for improving the design of one or more key components based on their impact on the performance of machine tool 105.

[0047] Maintenance module 195 is configured to determine maintenance activities to be performed on machine tool 105 based on performance. Maintenance module 195 is further configured to schedule maintenance activities based on their impact on performance.

[0048] The notification module 197 is configured to generate notifications indicating optimized control parameters and performance improvements derived from the optimized parameters, based on the output from the optimization module. Furthermore, the notification is also configured to generate notifications indicating recommendations generated by the recommendation module 192. Additionally, the recommendation module 192 is configured to generate notifications indicating scheduled maintenance activities for the machine tool 105, as well as periodic reminders for these maintenance activities.

[0049] Figure 2 A block diagram of a test setup 200 for constructing a digital twin of a machine tool 205 according to an embodiment of the present invention is shown. In this embodiment, the machine tool 205 is a CNC lathe. The CNC lathe is a three-axis machine including a spindle attached to a tool holder. The tool holder holds a cutting tool required for machining the workpiece. Similarly, the positioning of the workpiece relative to the cutting tool along the axis is achieved by ball screws and linear motion (LM) guides.

[0050] Test setup 200 includes a device 210 (similar to device 110) communicatively coupled to a controller 215 of machine tool 205. The controller 215 is configured to provide real-time status data to device 210, as previously used... Figure 1A As explained. Device 210 is further communicatively coupled to at least one first sensing unit 220A and at least one second sensing unit 220B. In one embodiment, the first sensing unit 220A and the second sensing unit 220B may be communicatively coupled to device 210 via controller 215.

[0051] The first sensing unit 220A includes a vibration sensor configured to measure vibrations associated with at least one component 225 on the structure of the machine tool 205. The at least one component 225 may be, for example, a bed or column associated with the machine tool 205. Those skilled in the art will understand that axis motors are not limited to external servo motors used with ball screws, but may also include other types of motors, such as linear motors for machine tools that do not employ ball screws. The vibration sensor may include, but is not limited to, capacitive sensors, piezoelectric sensors, and accelerometers. For example, the first sensing unit 220A may include an accelerometer attached to at least one of the at least one component 225. The second sensing unit 220B includes a temperature sensor configured to measure the temperature associated with at least one component 225 of the CNC lathe. The temperature sensor may include contact sensors (such as thermocouples and thermistors) and non-contact sensors (such as infrared sensors). For example, the second sensing unit 220B may include a thermocouple attached to at least one component 225. The first sensing unit 220A and the second sensing unit 220B are configured to provide the controller 215 with vibration and temperature measurements in real time, respectively.

[0052] The digital twin can be based on metadata associated with machine tool 205, historical data associated with machine tool 205, and a model of machine tool 205. Metadata may include specifications provided by the OEM of machine tool 205, which determine the type and maximum capacity of machine tool 205. Non-limiting examples of CNC lathe specifications include dimensions, center height, center distance, through-bed swing, sliding stroke, spindle bore, spindle nose, taper in nose, metric thread pitch, spindle power and torque, shaft power, and motor ratings. Metadata further includes physical and mechanical properties associated with different components of the CNC lathe. Non-limiting examples of these properties include dimensions, conductivity, elasticity, density, coefficient of thermal expansion, yield strength, tensile strength, fatigue strength, shear strength, and toughness. Historical data may include historical information related to the performance, maintenance, and health status of machine tool 205. The model of machine tool 205 is a hybrid model based on at least one physics-based model and at least one machine learning model associated with machine tool 205. This hybrid model correlates real-time status data from controller 215 with the real-time state of machine tool 205 based on metadata and historical data. In other words, the output of at least one machine learning model and at least one physics-based model at any given instance indicates the real-time state of machine tool 205. In this embodiment, the hybrid model of machine tool 205 includes one or more machine learning models and one or more physics-based models. The one or more machine learning models include a first machine learning model and a second machine learning model. The first machine learning model and the second machine learning model are trained using simulation data generated by a dynamic simulation model and a thermal simulation model, respectively.

[0053] The dynamic simulation model comprises multiple interconnected simulation models to simulate the dynamic properties of machine tool 205. In this embodiment, the dynamic simulation model includes a 3D simulation model and a 1D simulation model. The 3D simulation model is used to model the contact physics between components of machine tool 205, while the 1D simulation model is used to model the components of machine tool 205. The 1D simulation model includes at least one of a mass-spring-damper model and a lumped mass model corresponding to the components of machine tool 205. The dynamic simulation model is further configured based on condition data from controller 215.

[0054] The configured dynamic simulation model is executed via co-simulation in a suitable simulation environment to generate simulated values ​​corresponding to vibrations in at least one component 225 of machine tool 205. For example, the simulated values ​​are generated as vibration profiles for predefined time intervals corresponding to at least one component 225. Similarly, the simulated vibration values ​​are generated for different operating conditions of machine tool 205. The simulated vibration values ​​are further used to train a first machine learning model using a supervised learning algorithm to predict vibration values ​​based on condition data. However, it must be understood that the first machine learning model can be trained using other techniques, including but not limited to unsupervised learning algorithms, deep learning algorithms, and reinforcement learning.

[0055] The trained first machine learning model is further calibrated based on actual vibration values ​​measured by a first sensing unit 220A attached to at least one component 225. If the simulated vibration values ​​deviate from the actual vibration values ​​of the machine tool 205, the first machine learning model is retrained until the vibration values ​​predicted by the first machine learning model replicate the vibration values ​​measured by the first sensing unit 220A. The vibration values ​​correspond to the dynamic response of the machine tool 205. Based on the vibration values, a vibration spectrum can be generated using, for example, a Fast Fourier Transform (FFT). In other words, the first machine learning model is tuned to replicate a dynamic response substantially similar to that of the machine tool 205 for any given operating condition.

[0056] Similarly, the thermal simulation model is constructed using a finite element simulation model associated with machine tool 205. Furthermore, the thermal simulation model is configured based on real-time condition data. The configured thermal simulation model is executed in a suitable simulation environment to generate simulated values ​​corresponding to temperature values ​​associated with at least one component 225. For example, the simulated values ​​are generated as thermal profiles corresponding to at least one component 225 within a predefined time interval. Similarly, the simulated temperature values ​​are generated for different operating conditions. The simulated temperature values ​​are further used to train a second machine learning model using a supervised learning algorithm to predict the temperature in at least one component 225 based on condition data from controller 215. The second machine learning model is further calibrated based on actual temperature values ​​received from a second sensing unit 220B attached to at least one component 225. In other words, the second machine learning model is adjusted to replicate the thermal response of at least one component 225 based on any given operating condition.

[0057] In addition to the above, one or more machine learning models include a third machine learning model trained to identify one or more defects in at least one component 225. Defects include, but are not limited to, elongated wear, notch wear, pitting wear, plastic deformation, thermal cracking, fragmentation, and fatigue fracture. The third machine learning model is trained by artificially introducing each defect onto at least one component 225. Furthermore, a vibration spectrum corresponding to each defect is generated based on the output from the first machine learning model. Additionally, at least one data analysis technique is used to extract one or more features from the vibration spectrum indicating degradation associated with machine tool 205. These features can be further used to calculate a degradation index representing the level of degradation associated with at least one component 225. In one embodiment, spectral kurtosis associated with the vibration spectrum can be measured. Based on the spectral kurtosis, noise can be filtered out from the vibration spectrum using a suitable bandpass filter. Furthermore, harmonics present in the vibration spectrum can be analyzed to detect the presence of defective frequencies, i.e., frequencies indicating a specific defect. Furthermore, peak values ​​corresponding to the defective frequencies are analyzed to determine the degradation state due to the defect. For example, an increase in peak amplitude can indicate ongoing degradation. Similarly, the vibration spectrum can be subjected to a wavelength transformation function to detect the presence of non-stationary shock pulses caused by defects. For example, the wavelength transformation function may include one of discrete and continuous wavelength transformations. Alternatively, or in addition to the above, for certain defects, the degradation index may be indicated by features such as spectral energy, root mean square (RMS) velocity, envelope RMS velocity, crest factor, modified crest factor, or combinations thereof associated with the vibration spectrum. Furthermore, a third machine learning model is trained based on the extracted features corresponding to each defect. In one example, the third machine learning model may be trained based on a supervised learning algorithm such as the k-NN algorithm. After training, the third machine learning model can predict defects in machine tool 205 based on features of the vibration spectrum associated with at least one component 225.

[0058] Figure 3 A flowchart is depicted of an exemplary method 300 for instantaneous performance management of a machine tool according to an embodiment of the present invention.

[0059] At step 305, real-time status data associated with the machine tool is received from one or more sources. For example, the real-time status data corresponds to the machine tool's operating status received in real-time from the machine tool's controller. The status data may correspond to, for example, the type of machining operation, the type of cutting tool, sensor data, feed rate, cutting speed, spindle power, axis torque, axis speed, spindle speed, strokes per minute for each axis, and stroke range. Here, sensor data includes time-series data corresponding to vibration and temperature received in real-time from sensing units associated with the machine tool. Furthermore, the status data also includes specifications associated with the machine tool and material test data associated with the workpiece. Based on the specifications and control parameters, such as cutting speed, feed rate, and depth of cut, the maximum performance of the machine tool is calculated. For example, maximum performance indicates the ideal cycle time required to complete machining on the workpiece. In another example, maximum performance is machining accuracy.

[0060] At step 310, at least one parameter value that may affect the performance of the machine tool is calculated based on condition data. In this embodiment, the at least one parameter value corresponds to the dynamic stiffness associated with one or more critical components of the machine tool. One or more critical components are selected such that the behavior of these components affects the performance of the machine tool. In a preferred embodiment, the one or more critical components include a spindle, LM guide, ball screw, and cutting tool. However, it must be understood that the one or more critical components may further include other components, including but not limited to bearings, couplers, and motors associated with shafts and spindles. In another embodiment, the at least one parameter value includes both static stiffness and dynamic stiffness. More specifically, a physics-based model is used to calculate the stiffness values ​​of one or more critical components based on operating conditions and specifications.

[0061] Here, those skilled in the art must understand that the sensor data present within the real-time condition data is an indication of changes in stiffness values. Generally, stiffness values ​​deteriorate with increasing temperature. Similarly, increased vibration indicates increased displacement of the component. Increased displacement is yet another indicator of decreased stiffness values. Therefore, changes in stiffness values ​​can affect machining accuracy.

[0062] The physics-based model may include one or more of the following: a mathematical model, a reduced-order model, or a finite element (FE) model corresponding to one or more key components. In one embodiment, the physics-based model is first configured based on real-time status data from controller 115. Furthermore, stiffness values ​​for one or more key components are determined by finite element analysis (FEA) simulation using the configured physics-based model. For example, the results of the FEA simulation may indicate the stiffness value associated with each component. Similarly, stiffness values ​​are calculated using predefined time intervals (e.g., every second). In another embodiment, dynamic stiffness may be calculated using predefined mathematical models from sensor data generated by one or more sensing units.

[0063] At step 315, the digital twin of the machine tool is configured based on parameter values. In this embodiment, the digital twin is configured by updating based on condition data and stiffness values. More specifically, the metadata of the digital twin is updated based on specifications, and the hybrid model is updated based on condition data. The configured digital twin represents the real-time state of the machine tool.

[0064] At step 320, the behavior of the component is simulated in a simulation environment based on the configured digital twin. For example, this behavior can be related to the stability of a machine tool. To simulate this behavior, a simulation instance is first generated based on the configured digital twin. The simulation instance represents the machine-readable state of the machine tool corresponding to the configured digital twin. Similarly, multiple simulation instances corresponding to multiple operating conditions can be generated. In this example, these multiple operating conditions can correspond to state data at different times. Furthermore, the simulation instance is executed in the simulation environment using a simulation model. Here, the simulation model correlates the machine's state with its performance. In this embodiment, the machine's performance is measured based on cycle time. The simulation environment can be provided by different computer-aided simulation tools. In this embodiment, the simulation instance is executed through co-simulation of a 1D simulation model and a 3D simulation model. After execution, simulation results corresponding to each simulation instance are generated. The simulation results indicate the performance of the machine tool corresponding to each simulation instance.

[0065] At step 325, the impact on machine tool performance is predicted based on the behavior of the components in the simulation environment. In one embodiment, the impact on machine tool performance is determined based on the generated simulation results. For example, the simulation results may include a harmonic spectrum. The harmonic spectrum can indicate the presence of harmonics at a cutting pass frequency associated with the cutting tool. The cutting pass frequency is defined as the frequency at which the teeth of the cutting tool pass through a specific point during rotation. The cutting pass frequency is determined based on the angular velocity of the cutting tool and the number of teeth on the cutting tool, as determined by simulation. Furthermore, the harmonics can be further analyzed to determine characteristics indicating a stability index. The stability index can be defined as a numerical value indicating the stability level of the machine tool. For example, the stability index can be calculated based on the amplitude of the harmonics. In another example, the simulation results may include vibration waveforms associated with one or more key components in a time series. The vibration waveforms are analyzed to determine characteristics indicating a stability index. The characteristics may include, but are not limited to, spectral energy, spectral kurtosis, root mean square (RMS) velocity, envelope RMS velocity, crest factor, and modified crest factor. Predictive models can be used to correlate these characteristics with the stability index. The stability index can be further correlated to the accuracy of the machine tool using predefined mathematical relationships. In another example, this effect can be predicted, for example, as the cycle time corresponding to different spindle speeds. Generally, a lower stability index may lead to a higher cycle time, and a higher stability index may lead to a lower cycle time.

[0066] In another embodiment, at least one feature of the harmonic spectrum or vibration waveform can be used to identify defects in one or more components. Furthermore, said features can be correlated to a degradation index based on a predictive model. The degradation index can be correlated to the remaining service life of the machine tool. In one embodiment, simulation-assisted failure mode and effects analysis (FMEA) can be performed to determine the remaining service life of one or more critical components based on one or more defects. The time to failure of the machine tool is determined based on the RUL (Remaining Usage Limit). For example, the time to failure is equal to the shortest RUL value among the RUL values ​​of one or more critical components.

[0067] In step 330, the operation of the machine tool is optimized based on its impact on the machine tool's performance, as shown in the following reference. Figure 4 The explanation given.

[0068] Figure 4 A flowchart depicts an exemplary method 400 for optimizing machine tool operation according to an embodiment of the present invention.

[0069] At step 405, at least one control parameter for improving the machine tool's performance is identified based on the predicted impact on machine tool performance. The control parameter may include one or more of cutting speed, feed rate, and depth of cut.

[0070] At step 410, the values ​​of the control parameters are calculated based on the predicted impact on machine tool performance. In one embodiment, an optimization algorithm is used to calculate the values ​​of the control parameters. In one embodiment, the optimization algorithm is a multivariate optimization algorithm that optimizes multiple control parameters to minimize the impact on performance.

[0071] In step 415, the operation of the machine tool is simulated based on the calculated values ​​of the control parameters. For example, the operation of the machine tool can be simulated by generating a simulation instance based on the calculated values ​​of the control parameters.

[0072] At step 420, the simulation performance of the machine tool from the simulated operation is compared to a threshold. The threshold can be predefined by the machine tool operator. For example, the operator can specify that the cycle time cannot exceed 10 minutes. If the simulation performance is less than the threshold, step 410 is repeated to fine-tune the control parameters. Otherwise, step 425 is executed.

[0073] In step 425, the calculated values ​​of the control parameters are applied to operate the machine tool. More specifically, the calculated values ​​of the control parameters are set on the machine tool's controller.

[0074] In one embodiment, device 110 can generate recommendations for improving the design of one or more critical components based on their impact on machine tool performance. In one embodiment, the recommendations can be generated based on information stored in a knowledge graph. The knowledge graph can be constructed by the OEM based on, for example, the configuration or replacement history of multiple machine tools. The knowledge graph can be maintained on device 110 or on a different system (not shown) associated with the OEM communicatively coupled to device 110. A graphical query language can be used to query the knowledge graph. The query can be based on keywords corresponding to the impact on machine tool performance, defects, part numbers of defective components, etc. Based on the query, one or more responses are returned from the knowledge graph. The one or more responses correspond to recommendations for improving the design of one or more critical components. For example, if a spindle with a particular part number is identified as frequently defective, a recommendation can be associated with replacing that spindle with another spindle of a different part number.

[0075] In another embodiment, maintenance activities for a machine tool can be scheduled based on its impact on the machine tool's performance. Maintenance activities can be scheduled based on analysis of the machine tool's performance and on preferences predefined by the machine tool's operator. For example, maintenance activities can be scheduled if the machine tool's performance degrades frequently. The frequency of degradation can be determined based on the number of times the machine tool's operation has been optimized. Maintenance activities can be scheduled, for example, after five instances of performance degradation. Furthermore, the operator can be notified of the maintenance activity schedule, for example, on the human-machine interface (HMI) associated with the machine tool. Additionally, periodic warning messages can be displayed on the HMI to prompt the execution of maintenance activities.

[0076] In one embodiment, machining activities are planned based on machine tool performance. For example, an operator may assign multiple machine tools to machine a workpiece. For instance, the first machine tool could be a CNC milling machine, the second a CNC grinding machine, and so on. After assigning the machine tools, the operator can specify control parameters for the machining operation on each machine tool, such as cutting speed, feed rate, and depth of cut. Furthermore, the operator can specify the cycle time for completing the machining operation. Based on the control parameters specified by the operator and historical values ​​of sensor data associated with the machine tool, stiffness values ​​associated with one or more key components of the machine tool are predicted. In one example, a digital twin of the machine tool can be configured based on the control parameters provided by the operator and historical values ​​of sensor data. For example, sensor data from the most recent operation of the machine tool can be considered for historical values. Furthermore, the behavior of the machine tool is simulated based on the stiffness values. The results from the simulation can be further used to determine whether the simulated performance of the machine tool meets the specified cycle time. If the simulation performance indicates a cycle time lower than the specified cycle time, a notification can include the cycle time associated with the simulation performance. Otherwise, if the simulation performance indicates a cycle time greater than the specified cycle time, a notification can be generated. For example, a notification could include a message such as "The machine tool is unable to meet the specified cycle time." In addition to the above, the machine tool can also generate one or more recommendations for achieving the specified cycle time.

[0077] Advantageously, the present invention helps to calculate the overall equipment efficiency (OEE) associated with a machine tool for each job. OEE is calculated using the following equation:

[0078] OEE=P*Q*A (1)

[0079] Where P is performance as a percentage, Q is quality as a percentage, and A is machine availability as a percentage. In the equation above, P is calculated using the present invention. Q is determined by the operator based on the quality of the finished workpieces. For example, if 9 out of a total of 10 finished workpieces meet the accuracy level specified by the machine, then the quality is 90%. A is determined based on, for example, the mean time between failures (MTBF) and mean time to repair (MTTR) associated with the machine tool. Specifically,

[0080] A = 100 * MTBF / (MTTR + MTBF) (2)

[0081] This invention is not limited to a specific computer system platform, processing unit, operating system, or network. One or more aspects of this invention can be distributed across one or more computer systems, such as servers configured to provide one or more services to one or more client computers or to perform a complete task in a distributed system. For example, one or more aspects of this invention can be implemented on a client-server system comprising components distributed across one or more server systems performing multiple functions according to various embodiments. These components include, for example, executable code, intermediate code, or interpreted code that communicate over a network using communication protocols. This invention is not limited to being executable on any particular system or group of systems, nor is it limited to any particular distributed architecture, network, or communication protocol.

[0082] While the invention has been described and illustrated in detail with the aid of preferred embodiments, it is not limited to the disclosed examples. Other modifications can be deduced by those skilled in the art without departing from the scope of the claimed invention.

Claims

1. A computer-based method for instantaneous performance management of a machine tool (105), the method comprising: The processing unit (155) receives real-time status data associated with one or more components (225) of the machine tool (105) from one or more sources (115), wherein the status data indicates one or more real-time operating conditions of the machine tool (105); Based on the condition data, a physics-based model is used to calculate dynamic stiffness values ​​associated with one or more key components that may affect the performance of the machine tool (105). The digital twin of the machine tool (105) is configured based on the dynamic stiffness value; Simulate the behavior of one or more key components in a simulation environment based on the configured digital twin; The impact on the performance of the machine tool (105) is predicted based on the behavior of one or more key components in the simulation environment; as well as The operation of the machine tool (105) is optimized based on its impact on the performance of the machine tool (105).

2. The method according to claim 1, wherein configuring the digital twin of the machine tool (105) based on the dynamic stiffness value comprises: The digital twin of the machine tool (105) is updated based on the condition data and the calculated dynamic stiffness value.

3. The method of claim 1, wherein simulating the behavior of the one or more key components in a simulation environment based on the configured digital twin includes: Simulation instances are generated based on the configured digital twin; as well as The simulation instance is executed in a simulation environment using a simulation model to generate simulation results that indicate the behavior of the one or more key components.

4. The method according to any one of claims 1 and 3, wherein the behavior of said one or more key components is associated with the stability of said machine tool (105).

5. The method according to any one of claims 1 and 3, wherein predicting the impact on the performance of the machine tool (105) based on the behavior of the one or more key components comprises: The impact on the cycle time associated with the machine tool (105) is predicted based on the stability of the machine tool (105).

6. The method according to any one of claims 1 and 3, wherein predicting the impact on the performance of the machine tool (105) based on the behavior of the one or more key components comprises: The impact on machining accuracy associated with the machine tool (105) is predicted based on the stability of the machine tool (105).

7. The method of any one of claims 1 and 3, wherein the behavior of the one or more critical components is associated with one or more defects in the one or more critical components.

8. The method according to any one of claims 1 and 3, wherein predicting the impact on the performance of the machine tool (105) based on the behavior of the one or more key components comprises: The remaining useful life of the one or more critical components is calculated based on one or more defects.

9. The method of claim 1, wherein optimizing the operation of the machine tool (105) based on its impact on the performance of the machine tool (105) comprises: Based on the aforementioned effects, at least one control parameter is identified for improving the performance of the machine tool (105); The values ​​of the control parameters are calculated based on their impact on the performance of the machine tool (105); The operation of the machine tool (105) is simulated based on the calculated values ​​of the control parameters; The simulation performance of the machine tool (105) from the simulated operation is compared with a threshold. as well as If the simulation performance of the machine tool (105) is greater than the threshold, then the calculated value of the control parameter is applied to operate the machine tool (105).

10. The method of claim 1, further comprising: One or more recommendations are generated based on the impact on the performance of the machine tool (105) to improve the design of one or more key components.

11. The method of claim 1, further comprising: Maintenance activities for the machine tool (105) are scheduled based on its impact on performance.

12. A device (110) for instantaneous performance management of a machine tool (105), the device (110) comprising: One or more processing units (155); as well as A memory unit (170) communicatively coupled to the one or more processing units (155), wherein the memory unit (170) includes a status management module (175) stored in the form of machine-readable instructions executable by the one or more processing units (155), wherein the status management module (175) is configured to perform the method steps according to any one of claims 1 to 11.

13. A system (100) for instantaneous performance management of a machine tool (105), the system (100) comprising: One or more sources (115) are configured to provide real-time status data associated with the machine tool (105); as well as The apparatus (110) according to claim 12 is communicatively coupled to the one or more sources (115), wherein the apparatus (110) is configured to perform instantaneous performance management of the machine tool (105) based on the real-time status data according to any one of method claims 1 to 11.

14. A computer program product wherein machine-readable instructions are stored, which, when executed by one or more processing units (155), cause the processing unit (155) to perform the method according to any one of claims 1 to 11.

Images