Adaptive machining parameter optimization method and system for numerical control machine tool

Abstract

The application relates to the technical field of machine tool machining parameter optimization, and provides an adaptive machining parameter optimization method and system for a numerical control machine tool.The method comprises the following steps: performing wear detection on a numerical control machine tool cutter to generate a wear state vector; reading a task, searching a machining database, and matching a corresponding first machining parameter set; predicting a state time sequence based on the wear state and the first parameter set; controlling the numerical control machine tool to execute the task and monitoring the state in real time by using the first parameter set; comparing the predicted time sequence with the monitored state to identify a first deviation point; judging whether the deviation exceeds a threshold value; if not, predicting the continuity of the deviation from the point as a starting point; and optimizing the machining parameter according to the result. The application solves the technical problem that the existing numerical control machine tool cannot dynamically adjust the machining parameter according to the cutter wear state during the machining process, which leads to the decline of machining precision and the loss of machining efficiency, and achieves the technical effects of improving the machining efficiency, reducing the cutter loss, ensuring the stability of machining quality, and the like by monitoring the cutter wear in real time and optimizing the machining parameter.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of machine tool processing parameter optimization, in particular to an adaptive processing parameter optimization method and system for a numerical control machine tool. BACKGROUND

[0002] With the continuous development of manufacturing industry and the progress of technology, numerical control machine tools are increasingly widely used in the field of precision machining. Numerical control machine tools realize efficient machining of complex parts through automatic control systems, and play an indispensable role in mass production. However, in the long-term use of numerical control machine tools, tool wear, changes in processing conditions and fluctuations in processing environment will have a significant impact on processing quality, processing efficiency and tool service life. The traditional setting of processing parameters of numerical control machine tools mainly depends on preset processing specifications and experience parameters, and it is difficult to realize real-time adaptation and intelligent optimization. Especially, tool wear will inevitably occur during machining, and the decay of its state directly leads to the drift of key physical quantities such as cutting force, vibration and temperature. If the parameters are not adjusted in time, it may cause product quality problems such as workpiece size out-of-tolerance and surface roughness decline, or even tool collapse and machine tool damage, resulting in unplanned downtime and economic losses. SUMMARY

[0003] The present application provides an adaptive processing parameter optimization method and system for a numerical control machine tool, aiming to solve the technical problem that the existing numerical control machine tool cannot dynamically adjust the processing parameters according to the tool wear state during the machining process, resulting in a decrease in machining precision and a loss of processing efficiency.

[0004] The first aspect of the present application provides an adaptive processing parameter optimization method for a numerical control machine tool, the method comprising: detecting the wear state of the tool of the numerical control machine tool to generate a tool wear state vector; reading a to-be-machined task, retrieving a machining database of the same type of machining task, and matching a first set of processing parameters corresponding to the tool wear state vector in the machining database; based on the tool wear state vector and the first set of processing parameters, predicting the state during the machining process to generate a state prediction time sequence; controlling the numerical control machine tool to execute the to-be-machined task with the first set of processing parameters, and simultaneously monitoring the machining state in real time, comparing with the state prediction time sequence, and identifying a first state deviation point; judging whether the deviation state of the first state deviation point exceeds a preset threshold, if not, performing a continuity prediction of the state deviation starting from the first state deviation point, and optimizing the processing parameters according to the continuity prediction result.

[0005] In another aspect of the present disclosure, an adaptive machining parameter optimization system for a numerical control machine tool is provided, which comprises: a wear detection module configured to detect a wear state of a tool of the numerical control machine tool and generate a tool wear state vector; a machining parameter matching module configured to read a machining task to be processed, search a machining database of the same type of machining task, and match a first set of machining parameters corresponding to the tool wear state vector in the machining database; a state prediction module configured to perform state prediction during machining based on the tool wear state vector and the first set of machining parameters and generate a state prediction time sequence; a state monitoring module configured to control the numerical control machine tool to perform the machining task to be processed using the first set of machining parameters and simultaneously perform real-time monitoring of a machining state, compare the machining state with the state prediction time sequence, and identify a first state deviation point; and a machining parameter optimization module configured to determine whether a deviation state of the first state deviation point exceeds a preset threshold, and if not, perform continuous state deviation prediction starting from the first state deviation point and optimize machining parameters based on a result of the continuous state deviation prediction.

[0006] The one or more technical solutions provided in the present disclosure have at least the following technical effects or advantages:

[0007] The adaptive machining parameter optimization method for the numerical control machine tool first detects a wear state of a tool and generates a wear state vector. Then, a machining task to be processed is read and a database is queried to find a set of machining parameters corresponding to the wear state. Subsequently, the wear state and the set of machining parameters are used to predict state changes during machining and generate a prediction time sequence. Then, during machining, the set of machining parameters is used to control the machine tool to perform the task and the machining state is monitored in real time. At the same time, the actual state is compared with the prediction time sequence to identify a state deviation point. If the deviation point does not exceed a preset threshold, continuous state deviation prediction is performed starting from the deviation point. Finally, machining parameters are optimized based on the prediction result to ensure machining accuracy and efficiency.

[0008] The above description is only a summary of the technical solutions of the present disclosure. In order to more clearly understand the technical means of the present disclosure, the specific embodiments of the present disclosure can be implemented in accordance with the content of the description, and in order to make the above and other purposes, features and advantages of the present disclosure more obvious and easy to understand, the following specific embodiments of the present disclosure are described. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following will briefly introduce the drawings needed to be used in the embodiments description. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and other drawings can also be obtained by those skilled in the art without any creative effort based on these drawings.

[0010] Figure 1This is a flowchart illustrating an adaptive machining parameter optimization method for CNC machine tools in one embodiment.

[0011] Figure 2 This is a system architecture diagram for adaptive machining parameter optimization of CNC machine tools in one embodiment.

[0012] Explanation of reference numerals in the attached diagram: Wear detection module 11, machining parameter matching module 12, condition prediction module 13, condition monitoring module 14, machining parameter optimization module 15. Detailed Implementation

[0013] This application provides an adaptive machining parameter optimization method and system for CNC machine tools, which solves the technical problem that existing CNC machine tools cannot dynamically adjust machining parameters according to tool wear during machining, resulting in decreased machining accuracy and loss of machining efficiency.

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0015] It should be noted that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such process, method, product, or device.

[0016] Example 1, as Figure 1 As shown, this application provides an adaptive machining parameter optimization method for CNC machine tools, the method comprising:

[0017] The wear condition of the cutting tools on CNC machine tools is detected, and a tool wear condition vector is generated.

[0018] In the embodiments of this application, the specific content is described as follows:

[0019] In the method, first, vibration sensors, temperature sensors, acoustic sensors, position sensors, force sensors, etc. laid on the numerical control machine tool are used to collect the state of the tool in the machining process in real time, and monitor various physical signals generated by the tool in the machining process, such as vibration, temperature change, acoustic fluctuation, relative position change of the tool, cutting force, etc. The changes of these signals are often closely related to the degree of tool wear, so the wear state of the tool can be judged by collecting these signals. Subsequently, by extracting the time sequence characteristics of these signals, a set of characteristic values representing the tool wear state can be obtained. These characteristic values will be spliced according to the preset vector template to form a tool wear state vector. This tool wear state vector usually includes multiple dimensions of data, each dimension representing a different physical characteristic, such as vibration amplitude, temperature change amplitude, acoustic frequency, etc. The values of these dimensions reflect the working state of the tool in the machining process, and also provide an important basis for subsequent machining parameter optimization to ensure machining quality and efficiency, and reduce production interruptions caused by tool failure or excessive wear.

[0020] Read the machining task to be processed, retrieve the machining database of the same type of machining task, and match the first set of machining parameters corresponding to the tool wear state vector in the machining database.

[0021] In one embodiment, first, the current machining task to be processed is obtained, including the type of workpiece to be machined, the material of the workpiece, the expected machining accuracy, etc. The relevant parameters of the machining task to be processed can be obtained through user input or process dispatching. Subsequently, a machining database storing a large amount of historical machining task data is connected. This database contains the machining parameter sets of past machining tasks, and each set of machining tasks stores relevant machining parameters (such as cutting speed, feed speed, cutting depth, coolant consumption, etc.), as well as corresponding workpiece and tool state data. Then, according to the current machining task to be processed, the historical machining task data of the same type of workpiece is matched from the machining database, and the Euclidean distance is used to compare the current tool wear state vector with the tool state data in the historical machining task data, match the historical machining task data with the smallest similar distance, and extract the machining parameters in the historical machining task data as the first set of machining parameters corresponding to the tool wear state vector. This first set of machining parameters not only can predict the performance of the tool in the current machining task to some extent, but also can provide a basis for the next step of state prediction and optimization adjustment, ensuring the stability and precision of the machining process.

[0022] Based on the tool wear state vector and the first set of machining parameters, the state prediction in the machining process is performed to generate a state prediction time sequence.

[0023] In one embodiment, after obtaining the tool wear state vector and the first set of machining parameters, the tool wear state vector and the first set of machining parameters are input into the twin machine tool for simulation simulation, simulating various state changes that may occur during the machining process of the part and the tool, such as temperature rise, vibration increase, etc. By recording and organizing these simulation results, a state prediction time sequence of the machining process can be obtained, which details the state change trend of the part and the tool at each time and the change of the machining environment during the machining process, providing a key basis for subsequent real-time monitoring and machining parameter optimization, helping the system to make timely adjustments during the machining process, and ensuring the stability and precision of the machining process.

[0024] Further, the application provides state prediction during the machining process based on the tool wear state vector and the first set of machining parameters, and generates a state prediction time sequence, including:

[0025] The numerical control machine tool is modeled and a twin machine tool is constructed. The to-be-machined task is analyzed, and the structural attributes and material attributes of the to-be-machined part are read. After conditionally configuring the twin machine tool with the structural attributes, the material attributes, and the tool wear state vector, the first set of machining parameters is loaded, the state of the part and the tool during the machining process is read, and the state prediction time sequence is generated.

[0026] Preferably, before performing state prediction, a large amount of machine tool data (such as mechanical structure, motion mode, operation parameter, etc.) and basic principles of the machining process (such as cutting force, heat conduction, etc.) are obtained, and these data are input into the digital twin simulation software to construct a virtual twin machine tool. The twin machine tool can not only reflect the working state of the machine tool itself (such as position, vibration, temperature, etc.), but also simulate the changes of the tool, workpiece and machining process, which is used for subsequent precise real-time simulation of the actual machine tool in the virtual environment, so as to predict various problems in the workpiece machining process without actual machining. Subsequently, the machining task is analyzed to read the detailed information of the machining task, including the structural attributes and material attributes of the machined part, wherein the structural attributes include the geometric shape, size, geometric features, etc. of the part; the material attributes include the material, hardness, toughness, thermal conductivity, density, etc. of the part, which directly affect tool selection and machining method. Then, based on the obtained structural attributes, material attributes and tool wear state vector, the twin machine tool is conditionally configured, that is, these information is mapped into the twin machine tool to ensure that the virtual machine tool can simulate the actual working conditions faced by the part during machining, including the material, size, shape of the machined part and possible errors during machining, for example, according to the size and shape of the part, geometric adjustment is performed in the virtual environment, and the cutting force model and heat conduction model are adjusted according to the material attributes to reflect the machining behavior of different materials, and the current tool health state is reflected according to the tool wear state vector. Then, the first set of machining parameters is loaded into the configured twin machine tool, and the twin machine tool will start real-time simulation of the machining process according to the loaded first set of machining parameters. During the simulation process, the running state of the twin machine tool is continuously monitored, and the part and tool states during machining, such as vibration condition, temperature change, etc. are read and recorded in real time. Finally, the recorded part and tool state data are sorted according to the time stamp to form a state prediction time sequence in the machining process, which contains the trend of the key states of the tool and the part changing with time during the machining process, providing a basis for subsequent state monitoring and machining parameter adjustment, thereby improving the stability and precision of the machining process.

[0027] Further, the application provides that the states of the part and the tool include at least vibration state, temperature state, acoustic wave state and position state.

[0028] Optionally, the real-time recorded state data of the parts and tools at least includes vibration state, temperature state, acoustic wave state and position state, wherein the vibration state refers to mechanical vibration and its change generated between the tool and the part during the machining process, which can be caused by various factors such as tool wear, uneven cutting force, unstable workpiece clamping, etc. During the machining process, vibration not only affects the machining accuracy, but also can cause the tool to wear or damage prematurely. The temperature state reflects the thermal change of the tool and the part during the machining process. During the cutting process, the friction will cause the temperature of the tool and the workpiece surface to rise, and the excessively high temperature can affect the hardness and wear resistance of the tool, accelerate the wear of the tool, and also can affect the surface quality of the part, causing deformation or burn. The acoustic wave state refers to the acoustic wave signal generated during the machining process. Each material will emit a specific frequency sound during machining, and tool wear or abnormal state (such as excessive cutting, jamming, etc.) during the machining process will cause the change of the sound frequency. The position state refers to the position change of the tool and the part during the machining process. During the machining process, the position change of the tool directly determines the machining accuracy and the surface quality of the workpiece. Any small error in the position of the tool can cause shape error or size error of the workpiece, affecting the machining quality.

[0029] controlling the numerical control machine tool to perform the to-be-machined task according to the first set of machining parameters, and synchronously performing real-time monitoring of the machining state, comparing the real-time monitoring state with the state prediction time sequence, and identifying a first state deviation point.

[0030] In one embodiment, the first set of machining parameters is first transmitted to the numerical control machine tool, and the numerical control machine tool performs the to-be-machined task according to the parameters. During the machining process, various sensors installed on the machine tool are used to monitor the machining state in real time, including vibration, temperature change, acoustic wave fluctuation, position change, etc. By sorting these data according to the time stamp, a real-time monitoring time sequence is formed. Then, the obtained real-time monitoring time sequence is aligned with the state prediction time sequence, and the alignment result is compared to evaluate whether the machining process is consistent with the prediction result. If there is a deviation between the real-time monitoring time sequence and the state prediction time sequence, all data points with deviation are identified as the first state deviation point, which helps the system to discover potential risks in the machining process in time, avoids the generation of defective products, and provides data support for subsequent machining parameter optimization and adjustment.

[0031] Further, the application provides that the numerical control machine tool is controlled to perform the to-be-machined task according to the first set of machining parameters, and real-time monitoring of the machining state is synchronously performed, the real-time monitoring state is compared with the state prediction time sequence, and a first state deviation point is identified, including:

[0032] The sensor monitoring network is configured according to the state type contained in the state prediction time sequence; when the numerical control machine tool is controlled by the first set of machining parameters to perform the to-be-machined task, the sensor monitoring network is synchronously activated to perform real-time monitoring, and a real-time monitoring time sequence is generated; the real-time monitoring time sequence and the state prediction time sequence are aligned and compared according to the time sequence correspondence relationship, and the first state deviation point of state inconsistency is identified.

[0033] Preferably, various state types related to the machining process are extracted from the state prediction time sequence, such as vibration, temperature, sound wave, position, etc., which represent key physical characteristics that may affect machining quality and tool life during the machining process. According to these state types, corresponding sensor networks are configured, and according to the working area of the machine tool, the characteristics of the machining task and the specific position of the workpiece, these sensor networks are laid out, each sensor network at least including a vibration sensor, a temperature sensor, a sound wave sensor and a position sensor. Subsequently, the numerical control machine tool is controlled by the first set of machining parameters to perform the to-be-machined task, and the sensor monitoring network is synchronously activated to start real-time monitoring of various states during the machining process. During the monitoring process, the sensor network continuously collects various physical signal data of the tool and the part, including vibration, temperature, sound wave, displacement, etc., and transmits them back to the system. The system arranges these data in time sequence to form a real-time monitoring time sequence. Then, the real-time monitoring time sequence and the state prediction time sequence are aligned on the time axis to ensure that the real-time data and the prediction data are compared at the same time point. After time alignment, the real-time monitoring data at each time point is compared with the data in the state prediction time sequence one by one, and the difference between the two is calculated. If there is a deviation between the actual monitoring data and the prediction data at a certain time, it indicates that an abnormality or deviation may occur in the machining process. At this time, the data point corresponding to the time with deviation is taken as a first state deviation point, indicating that there is inconsistency between the real-time monitoring data and the state prediction time sequence. This ensures that the system can timely identify abnormal states that may occur during the machining process and respond quickly to improve machining quality.

[0034] It is judged whether the deviation state of the first state deviation point exceeds a preset threshold, and if not, a continuity prediction of state deviation is performed from the first state deviation point as a starting point, and machining parameter optimization is performed according to the continuity prediction result.

[0035] In one embodiment, for the identified first state deviation point, it is determined whether the deviation state of the state deviation point exceeds a corresponding preset threshold, for example, whether the deviation of the temperature exceeds a temperature deviation threshold. If the deviation state of the first state deviation point does not exceed the threshold, the first state deviation point is taken as a starting point for subsequent state deviation continuity prediction, that is, deviation propagation analysis is performed according to the deviation attribute and deviation value of the first state deviation point to predict the state change of the tool and the part in the subsequent machining process, and a continuity prediction result is generated. When the continuity prediction result exceeds the preset threshold, the machining parameters are automatically optimized, such as adjusting the cutting speed, the feed rate, the coolant flow rate, etc., to optimize the machining process, ensure the accuracy and efficiency of the machining, and prolong the service life of the tool.

[0036] Further, the application provides state deviation continuity prediction starting from the first state deviation point, and machining parameter optimization based on the continuity prediction result, including:

[0037] reading the first deviation attribute and the first deviation value of the first state deviation point; performing deviation propagation analysis based on state prediction timing according to the first deviation attribute and the first deviation value to generate the continuity prediction result; determining whether the continuity prediction result contains a predicted deviation point whose deviation state exceeds a preset threshold; if so, performing deviation correction at the first state deviation point based on the continuity prediction result to complete the machining parameter optimization.

[0038] Preferably, for the first state deviation point that does not exceed the preset threshold, the first deviation attribute and the first deviation value of the first state deviation point are read, wherein the first deviation attribute refers to the type of state that deviates, such as temperature deviation, sound wave deviation, vibration deviation, and position deviation; and the first deviation value refers to the size of the state type deviation, such as the difference between the actual temperature and the predicted temperature for temperature deviation, and the difference between the actual vibration amplitude and the predicted vibration amplitude for vibration deviation. Subsequently, the obtained first deviation attribute and first deviation value are input into the pre-constructed deviation transition analyzer for deviation transition analysis, in which process, the deviation transition analyzer predicts a possible actual state time sequence according to the received first deviation attribute and first deviation value, and aligns the actual state time sequence with the state prediction time sequence for the same deviation identification as described above to obtain a continuity prediction result, which shows that under the current deviation state, the subsequent deviation points that may occur in the machining process, such as continuous temperature rise and further vibration increase. Then, it is determined whether there is a predicted deviation point in the continuity prediction result that exceeds the preset threshold. If the deviation point in the prediction exceeds the preset threshold, deviation correction is performed based on the continuity prediction result, the purpose of which is to avoid unacceptable deviations in the machining process by adjusting the machining parameters, and to ensure the stability and accuracy of the machining process, for example, if the temperature is predicted to be too high, the cooling liquid flow rate can be adjusted, the cutting depth can be reduced, and the cutting speed can be reduced; if the vibration is predicted to be too large, the feed speed can be reduced, the tool path can be optimized, or a different tool type can be selected. After the machining parameters are optimized, the adjustment results are fed back in real time, and the machining state is continuously monitored. If a new deviation point appears, prediction and correction are continuously performed to ensure that the machining process is always in the best state.

[0039] Further, the application provides that the deviation transition analysis based on the state prediction time sequence is performed according to the first deviation attribute and the first deviation value to generate the continuity prediction result, comprising:

[0040] The historical deviation feature data containing the deviation attribute and the deviation value, and the historical state prediction time sequence and the historical actual state time sequence are collected; starting from the first historical deviation point in the historical state prediction time sequence, the state, action, transition, and reward mechanism are trained based on reinforcement learning according to the first deviation feature of the first historical deviation point in the historical deviation feature data and the deviation transition feature after the first historical deviation point in the historical actual state time sequence, and the deviation transition analyzer is constructed; the deviation transition analyzer is used to perform deviation transition analysis on the first deviation attribute, the first deviation value, and the state prediction time sequence, and the continuity prediction result is output.

[0041] Optionally, first, historical deviation feature data, historical state prediction time series and historical actual state time series are collected from historical logs, wherein the historical deviation feature data includes deviation attributes (such as temperature, vibration, sound wave, etc.) and corresponding deviation values that appear in the historical machining process, to reflect the difference between the actual state and the predicted state in the historical machining process; the historical state prediction time series provides the state expected value for each time point in the historical machining process; and the historical actual state time series records the actual state of the tool and the part in the machining process. Subsequently, from the first historical deviation point in the historical state prediction time series, the first deviation feature related to the first historical deviation point is extracted from the historical deviation feature data, and the deviation features after the first historical deviation point are taken as deviation transfer features. Then, in order to further accurately predict and analyze the transfer process of the deviation, the deviation transfer analyzer is trained through reinforcement learning, and the goal of the reinforcement learning is to optimize the deviation prediction process by learning the state change (state), the measures taken (action), the state transition (transition) and the result feedback (reward) in the historical data, wherein the state (State) refers to the current state in the machining process, such as temperature, vibration, sound wave, etc.; the action (Action) refers to the operation that the system can take, and the action affects the state change of the machining process; the transition (Transition) refers to the process of the system transferring from one state to another state; and the reward mechanism (Reward) refers to giving a certain reward or punishment according to the current state and the action taken, and the goal of the reward is to guide the system to take measures that can optimize the machining process and reduce the deviation, and the punishment is usually related to unqualified results or efficiency reduction in the machining process, and the reward is related to positive results such as deviation reduction and machining precision improvement. Through continuous iterative learning, the reinforcement learning can gradually build a deviation transfer analyzer that can accurately predict the development of the deviation state, and the deviation transfer analyzer can simulate the transfer process of the deviation state in the machining process and output a continuous prediction result according to the input first deviation attribute, first deviation value and state prediction time series, and the continuous prediction result shows the possible state changes in the machining process, including the aggravation or recovery of the deviation, and the system can adjust the machining parameters according to the predicted deviation trend to avoid the further development of the deviation, and ensure the stability and precision of the machining process.

[0042] Further, the present application provides that the state representation includes the first historical deviation point and other historical deviation points located after the first historical deviation point, the action is a transition operation from the first historical deviation point to the other historical deviation points, and the reward is a positive feedback when the transition relationship is correctly predicted.

[0043] Preferably, the state represents the first historical deviation point and other historical deviation points, reflecting different deviation states in the machining process, and by inputting these points as states, an accurate perception of the dynamic changes in the machining process can be established. The action represents the transition operation from the first historical deviation point to other historical deviation points, and each transition operation means a change in the state of the machining process, which may be an increase in vibration caused by temperature rise, or a decrease in machining precision caused by tool wear. The reward mechanism represents positive feedback when the transition relationship is correctly predicted. In the framework of reinforcement learning, the reward mechanism is used to evaluate whether the behavior of the system meets the expectations. The reward is a signal fed back by the system according to the current state and the action taken, indicating the effectiveness of the operation. When the state transition is correctly predicted and effective measures are taken, positive feedback will be obtained, which means that the current transition operation effectively reduces the machining deviation or optimizes the machining precision. If the transition relationship of the deviation is not correctly predicted or the measures taken fail to effectively avoid the deviation, negative feedback will be received, indicating that the prediction and adjustment strategy need to be improved. By continuously learning the transition relationship between the historical deviation points, the behavior strategy can be gradually optimized in the process of reinforcement learning. Whenever the deviation transition analyzer successfully predicts the deviation state, it will receive a reward, thereby improving the deviation transition analyzer's ability to predict the deviation of the machining process, and ultimately achieving adaptive optimization of the machining parameters, improving the machining precision and reducing tool wear.

[0044] Further, the application provides deviation correction at the first state deviation point based on the continuity prediction result, completing machining parameter optimization, including:

[0045] If the first deviation attribute only includes the temperature attribute, the adaptation and optimization of the cooling liquid parameters are performed; if the first deviation attribute includes multiple attributes, the first deviation attribute and the first deviation value are input into the pre-constructed cutting parameter optimization library for correction parameter matching.

[0046] Optionally, when the first deviation attribute is only temperature, it is judged whether the temperature change in the machining process exceeds the preset safety range. If it is detected that the temperature exceeds the safety range, the cooling liquid parameter is optimized according to the real-time temperature data, that is, the cooling effect can be enhanced by increasing the cooling liquid flow to reduce the temperature of the tool and the workpiece, thereby preventing the machining problem caused by overheating. If the first deviation attribute includes not only temperature but also other attributes (such as vibration, sound wave, position error, etc.), the first deviation attribute (such as temperature, vibration, cutting force, etc.) and the corresponding first deviation value (such as temperature rise amplitude, vibration amplitude, etc.) are input into the pre-constructed cutting parameter optimization library, which is a database containing a large amount of historical data and optimization experience. The most suitable optimization scheme for the current machining condition can be found from the cutting parameter optimization library through cosine similarity. According to the optimization scheme, the machining parameters such as cutting depth, feed speed, tool path, cutting speed, etc. are automatically adjusted to compensate for the unstable factors caused by deviation in the machining process, thereby ensuring the stability of the machining process.

[0047] Further, the application provides a method for judging whether the deviation state of the first state deviation point exceeds the preset threshold, comprising:

[0048] If the deviation state of the first state deviation point exceeds the preset threshold, a correction instruction is issued; based on the correction instruction, deviation correction is performed at the first state deviation point according to the first deviation attribute and the first deviation value, and machining parameter optimization is completed.

[0049] Optionally, when the deviation state of the first state deviation point exceeds the preset threshold, a correction instruction is issued to activate the subsequent correction process. At this time, the first deviation attribute is judged to determine the number of deviation attributes. If there is only one attribute, the corresponding attribute is adapted and optimized, for example, temperature deviation adjusts the cooling liquid parameter (such as flow rate and temperature), vibration deviation adjusts the tool parameter (such as feed speed, cutting depth, and tool replacement), and position deviation recalibrates the machine tool and adjusts the tool path. If there are multiple attributes, the first deviation attribute and the first deviation value of the first state deviation point are matched with the cutting parameter optimization library to determine the required optimization scheme, and the machining parameters are optimized according to the matched optimization scheme to ensure the stability, accuracy and efficiency of the machining process, thereby improving the production quality and tool service life.

[0050] In summary, the embodiments of the application have at least the following technical effects:

[0051] This application embodiment first detects the wear state of the CNC machine tool's cutting tool, generating a cutting tool wear state vector. Then, it reads the task to be processed, searches a processing database of similar tasks, and matches a first processing parameter set corresponding to the cutting tool wear state vector in the database. Next, based on the cutting tool wear state vector and the first processing parameter set, it predicts the state during processing, generating a state prediction timeline. Then, it controls the CNC machine tool to execute the task to be processed using the first processing parameter set, and simultaneously monitors the processing state in real time, comparing it with the state prediction timeline to identify a first state deviation point. Finally, it determines whether the deviation of the first state deviation point exceeds a preset threshold. If not, it predicts the continuity of the state deviation starting from the first state deviation point, and optimizes the processing parameters based on the continuous prediction results. These technical effects collectively solve the technical problem that existing CNC machine tools cannot dynamically adjust processing parameters according to the cutting tool wear state during processing, leading to decreased processing accuracy and reduced processing efficiency. This achieves the technical effect of improving processing efficiency, reducing cutting tool wear, and ensuring stable processing quality by real-time monitoring of cutting tool wear and optimizing processing parameters.

[0052] Example 2, based on the same inventive concept as the adaptive machining parameter optimization method for CNC machine tools in the foregoing examples, such as... Figure 2 As shown, this application provides an adaptive machining parameter optimization system for CNC machine tools. The system includes: a wear detection module 11 for detecting the wear state of the cutting tool of the CNC machine tool and generating a cutting tool wear state vector; a machining parameter matching module 12 for reading the task to be machined, searching a machining database of similar machining tasks, and matching a first machining parameter set corresponding to the cutting tool wear state vector in the machining database; a state prediction module 13 for predicting the state during the machining process based on the cutting tool wear state vector and the first machining parameter set, and generating a state prediction time sequence; a state monitoring module 14 for controlling the CNC machine tool to execute the task to be machined with the first machining parameter set, and simultaneously monitoring the machining state in real time, comparing it with the state prediction time sequence, and identifying a first state deviation point; and a machining parameter optimization module 15 for determining whether the deviation state of the first state deviation point exceeds a preset threshold. If not, it performs a continuous prediction of the state deviation starting from the first state deviation point and optimizes the machining parameters based on the continuous prediction result.

[0053] Furthermore, the state prediction module 13 is also used to perform the following method:

[0054] The numerical control machine tool is twin modeling, and a twin machine tool is constructed; the machining task is analyzed, and structural attributes and material attributes of the part to be machined are read; after the twin machine tool is conditionally configured with the structural attributes, the material attributes and the tool wear state vector, the first set of machining parameters is loaded, the states of the part and the tool in the machining process are read, and the state prediction time sequence is generated.

[0055] Further, the state prediction module 13 is further used to execute the following method:

[0056] The states of the part and the tool at least include vibration states, temperature states, acoustic wave states and position states.

[0057] Further, the state monitoring module 14 is further used to execute the following method:

[0058] According to the state type configured by the state prediction time sequence, a sensing monitoring network is configured; when the numerical control machine tool executes the machining task controlled by the first set of machining parameters, the sensing monitoring network is synchronously activated to perform real-time monitoring, and a real-time monitoring time sequence is generated; the real-time monitoring time sequence and the state prediction time sequence are aligned and compared according to the time sequence correspondence, and the first state deviation point with inconsistent states is identified.

[0059] Further, the machining parameter optimization module 15 is further used to execute the following method:

[0060] The first deviation attribute and the first deviation value of the first state deviation point are read; deviation transfer analysis based on the state prediction time sequence is performed according to the first deviation attribute and the first deviation value, and the continuity prediction result is generated; it is judged whether the continuity prediction result contains a predicted deviation point with a deviation state exceeding a preset threshold; if yes, deviation correction is performed at the first state deviation point based on the continuity prediction result, and machining parameter optimization is completed.

[0061] Further, the machining parameter optimization module 15 is further used to execute the following method:

[0062] The historical deviation feature data containing deviation attributes and deviation values, the historical state prediction time sequence and the historical actual state time sequence are collected; starting from a first historical deviation point in the historical state prediction time sequence, deviation transfer analysis is performed on the first deviation attribute, the first deviation value and the state prediction time sequence based on the first deviation feature of the first historical deviation point in the historical deviation feature data and the deviation transfer feature after the first historical deviation point in the historical actual state time sequence, and a state, action, transition and reward mechanism is trained based on reinforcement learning to construct a deviation transfer analyzer; the continuity prediction result is output by the deviation transfer analyzer.

[0063] Further, the processing parameter optimization module 15 is further configured to execute the following method:

[0064] The state representation includes a first historical deviation point and other historical deviation points after the first historical deviation point, the action is a transition operation from the first historical deviation point to the other historical deviation points, and the reward is positive feedback when the transition relationship is correctly predicted.

[0065] Further, the processing parameter optimization module 15 is further configured to execute the following method:

[0066] If the first deviation attribute only includes the temperature attribute, the adaptation optimization of the coolant parameter is performed; if the first deviation attribute includes multiple attributes, the first deviation attribute and the first deviation value are input into a pre-constructed cutting parameter optimization library for correction parameter matching.

[0067] Further, the processing parameter optimization module 15 is further configured to execute the following method:

[0068] If the deviation state of the first state deviation point exceeds a preset threshold, a correction instruction is issued; based on the correction instruction, deviation correction is performed at the first state deviation point according to the first deviation attribute and the first deviation value, and the processing parameter optimization is completed.

[0069] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes a specific embodiment of the present application. The processes depicted in the drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0070] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

[0071] The present application and the drawings are only exemplary descriptions of the present application, and are considered to cover any and all modifications, changes, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art can make various modifications and changes to the present application without departing from the scope of the present application. Thus, if these modifications and changes of the present application belong to the scope of the present application and its equivalent technology, the present application intends to include these modifications and changes.

Claims

1. A method for adaptive machining parameter optimization for a numerically controlled machine tool, characterized in that, The method comprises: detecting the wear state of a tool of a numerical control machine tool to generate a tool wear state vector; reading a to-be-processed task, searching a machining database of the same type of machining task, and matching a first set of machining parameters corresponding to the tool wear state vector in the machining database; performing state prediction during machining based on the tool wear state vector and the first set of machining parameters to generate a state prediction time sequence; controlling the numerical control machine tool to execute the to-be-processed task by using the first set of machining parameters, and simultaneously performing real-time monitoring of the machining state, comparing the state prediction time sequence, and identifying a first state deviation point; judging whether the deviation state of the first state deviation point exceeds a preset threshold, and if not, performing continuity prediction of the state deviation starting from the first state deviation point, and performing machining parameter optimization according to the continuity prediction result; wherein, controlling the numerical control machine tool to execute the to-be-processed task by using the first set of machining parameters, and simultaneously performing real-time monitoring of the machining state, comparing the state prediction time sequence, and identifying a first state deviation point, comprises: configuring a sensing monitoring network according to the state types contained in the state prediction time sequence; synchronously activating the sensing monitoring network to perform real-time monitoring and generate a real-time monitoring time sequence when the numerical control machine tool is controlled to execute the to-be-processed task by using the first set of machining parameters; aligning and comparing the real-time monitoring time sequence and the state prediction time sequence according to the time sequence correspondence relationship to identify the first state deviation point with inconsistent states; wherein, performing continuity prediction of the state deviation starting from the first state deviation point, and performing machining parameter optimization according to the continuity prediction result, comprises: reading a first deviation attribute and a first deviation value of the first state deviation point; performing deviation transfer analysis based on the state prediction time sequence according to the first deviation attribute and the first deviation value to generate the continuity prediction result; judging whether the continuity prediction result contains a predicted deviation point with a deviation state exceeding a preset threshold; if yes, performing deviation correction at the first state deviation point based on the continuity prediction result to complete machining parameter optimization; wherein, performing deviation transfer analysis based on the state prediction time sequence according to the first deviation attribute and the first deviation value to generate the continuity prediction result, comprises: collecting historical deviation feature data containing deviation attributes and deviation values, as well as historical state prediction time sequences and historical actual state time sequences; starting from a first historical deviation point in the historical state prediction time sequence, performing state, action, transition, and reward mechanism training based on reinforcement learning according to a first deviation feature of the first historical deviation point in the historical deviation feature data and deviation transfer features after the first historical deviation point in the historical actual state time sequence, to construct a deviation transition analyzer, wherein the state representation includes the first historical deviation point and other historical deviation nodes after the first historical deviation point, the action is a transition operation from the first historical deviation point to other historical deviation nodes, and the reward is positive feedback when the transition relationship is correctly predicted. The deviation transfer analyzer is used to perform deviation transfer analysis on the first deviation attribute, the first deviation value and the state prediction time sequence, and output the continuity prediction result.

2. The adaptive machining parameter optimization method for a CNC machine tool of claim 1, wherein, Based on the tool wear state vector and the first set of machining parameters, state prediction during machining is performed to generate a state prediction time sequence, including: Performing twin modeling on the numerical control machine tool to construct a twin machine tool; Analyzing the machining task to be processed to read the structural attributes and material attributes of the part to be processed; After conditionally configuring the twin machine tool with the structural attributes, the material attributes and the tool wear state vector, the first set of machining parameters is loaded to read the state of the part and the tool during machining, and the state prediction time sequence is generated.

3. The adaptive machining parameter optimization method for a CNC machine tool of claim 2, wherein, The state of the part and the tool includes at least vibration state, temperature state, acoustic wave state and position state.

4. The adaptive machining parameter optimization method for a CNC machine tool of claim 1, wherein, Based on the continuity prediction result, deviation correction is performed at the first state deviation point to complete machining parameter optimization, including: If the first deviation attribute only includes temperature attribute, adaptive optimization of cooling liquid parameters is performed; If the first deviation attribute includes multiple attributes, the first deviation attribute and the first deviation value are input into a pre-constructed cutting parameter optimization library for correction parameter matching.

5. The adaptive machining parameter optimization method for a CNC machine tool of claim 1, wherein, It is further determined whether the deviation state of the first state deviation point exceeds a preset threshold, including: If the deviation state of the first state deviation point exceeds the preset threshold, a correction instruction is issued; Based on the correction instruction, deviation correction is performed at the first state deviation point according to the first deviation attribute and the first deviation value to complete machining parameter optimization.

6. An adaptive machining parameter optimization system for a numerically controlled machine tool, characterized by, The system is used to perform the adaptive machining parameter optimization method for a numerical control machine tool according to any one of claims 1-5, including: A wear detection module for detecting the wear state of the tool of the numerical control machine tool to generate a tool wear state vector; A machining parameter matching module for reading the machining task to be processed, retrieving the machining database of the same type of machining task, and matching the first set of machining parameters corresponding to the tool wear state vector in the machining database; A state prediction module for performing state prediction during machining based on the tool wear state vector and the first set of machining parameters to generate a state prediction time sequence; A state monitoring module for controlling the numerical control machine tool to execute the machining task to be processed with the first set of machining parameters, and simultaneously performing real-time monitoring of the machining state, comparing with the state prediction time sequence, and identifying a first state deviation point; A machining parameter optimization module for determining whether the deviation state of the first state deviation point exceeds a preset threshold, and if not, performing continuity prediction of state deviation with the first state deviation point as the starting point, and performing machining parameter optimization according to the continuity prediction result.

Images