A high-speed lathe intelligent fault diagnosis method and system

Abstract

The application relates to the technical field of numerical control equipment operation monitoring, in particular to a high-speed lathe intelligent fault diagnosis method and system; the method comprises the following steps: collecting operation data of a high-speed lathe, and constructing a production observation channel and a diagnosis observation channel; collecting observation data sequences of the production observation channel and the diagnosis observation channel, calculating a diagnosability quota, and generating a semi-healthy lock identification; judging whether the high-speed lathe meets a self-setting condition according to the diagnosability quota and the semi-healthy lock identification; when the high-speed lathe does not meet the self-setting condition, entering a self-setting sandbox mode, and updating high-speed lathe control parameters; and generating fault diagnosis information based on a machine tool state identification and the observation data sequences of the diagnosis observation channel. The application realizes a high-speed lathe intelligent fault diagnosis method which can complete high-speed lathe control parameter self-setting and intelligent fault diagnosis without sacrificing diagnosis capability.

Description

Technical Field

[0001] This invention relates to the field of CNC equipment operation monitoring technology, specifically to an intelligent fault diagnosis method and system for high-speed lathes. Background Technology

[0002] As the requirements for machining cycle time and surface quality of products such as automotive parts and 3C precision components continue to increase, high-speed lathes generally adopt electric spindles, high-response feed servos, and multi-axis linkage control. In order to maintain dimensional accuracy and surface consistency under long-term continuous machining, frequent start-stop and high-load cutting, machine tool manufacturers and users have widely adopted functions such as servo parameter self-tuning, automatic spindle tuning, automatic thermal error compensation, pitch compensation, and self-learning feed optimization. At the same time, online monitoring and intelligent analysis systems based on data such as vibration, current, temperature, and spindle load are also deployed on the production site to assist team leaders and maintenance personnel in judging whether the machine tool needs to be repaired or key components need to be replaced. Even with long-term operation of high-speed lathes and the widespread use of self-tuning, self-learning, and other adaptive functions, there is still a disconnect between operating status, control parameters, and fault diagnosis results. Specifically: First, components such as spindle bearings and feed screws may exhibit slight abnormal noises, excessive temperature rise, and heavier cutting loads, but under the cover of thermal compensation and gain adjustment, part dimensions may still be within acceptable limits for a short period, and the system may consistently display "normal." The machine tool remains in this semi-healthy state for extended periods without being individually flagged or scheduled for maintenance. Second, even in the aforementioned semi-healthy state or with signs of abnormality, servo self-tuning or spindle auto-tuning is still triggered as usual. The control system directly updates control parameters and diagnostic baselines with data collected under deteriorating conditions, solidifying the faulty dynamic characteristics as a new "normal standard." Even if the machine tool continues to deteriorate, intelligent diagnosis struggles to identify and alert the user in a timely manner. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent fault diagnosis method and system for high-speed lathes, so as to solve the problem of the disconnect between operating status, control parameters and fault diagnosis results mentioned in the background art.

[0004] To achieve the above objectives, the technical solution of the present invention is: an intelligent fault diagnosis method for high-speed lathes, comprising: S1. Collect the operating data of the high-speed lathe and construct the production observation channel and the diagnostic observation channel; apply compensation control, filtering control and self-tuning control to the production observation channel, and shield compensation control and self-tuning control to the diagnostic observation channel; S2. Real-time acquisition of observation data sequences from production observation channels and diagnostic observation channels; calculation of deviation data, compensation control quantity, and self-tuning adjustment quantity between the two channels; calculation of diagnostic quota based on deviation data, compensation control quantity, and self-tuning adjustment quantity and generation of semi-health lock-in flag as a new machine tool status flag. Wherein, the diagnostic quota is the quota data of the remaining degree of diagnostic information of the high-speed lathe under the current control state; the semi-health lock flag is a status flag indicating that the high-speed lathe is in a state without alarm but in a degraded state and restricts control behavior; the machine tool status flag includes the diagnostic quota and the semi-health lock flag. S3. Read the diagnostic quota and semi-health lock flag, and determine whether the high-speed lathe meets the self-tuning conditions based on the diagnostic quota and semi-health lock flag; perform normal self-tuning operation when the high-speed lathe meets the self-tuning conditions, and enter the self-tuning sandbox mode and update the high-speed lathe control parameters when the high-speed lathe does not meet the self-tuning conditions. The self-tuning sandbox mode is a self-tuning mode that performs self-tuning calculations within a limited operating range isolated from normal processing tasks, and does not directly modify the formal control parameters of the control system. S4. Based on the machine tool status identifier and the observation data sequence of the diagnostic observation channel, generate fault diagnosis information, and combine it with the updated high-speed lathe control parameters to perform control and handling operations on the high-speed lathe.

[0005] Preferably, in S1, the production observation channel and the diagnostic observation channel are two observation paths formed by splitting the high-speed lathe operation data based on a unified sampling time reference. The production observation channel is used to output the production observation data sequence under the conditions of performing compensation control, filtering control and self-tuning control, and the diagnostic observation channel is used to output the diagnostic observation data sequence under the conditions of performing signal preprocessing operation. The production observation channel and the diagnostic observation channel output the observation data sequence with the unified sampling time reference as the index.

[0006] Preferably, in step S2, the diagnostic quota is quota data calculated based on the deviation data, compensation control quantity, and self-tuning adjustment quantity of the production observation channel and the diagnostic observation channel within a preset statistical time window. This quota is used to characterize the degree of remaining diagnostic information of the high-speed lathe under the current control state. The data structure of the diagnostic quota includes a deviation quota field, a compensation quota field, and a self-tuning quota field. The diagnostic quota is constructed as follows: in each statistical time window, the deviation data, compensation control quantity, and self-tuning adjustment quantity are calculated respectively, and the deviation quota field, compensation quota field, and self-tuning quota field are obtained by combining them according to preset weight coefficients. The deviation quota field, compensation quota field, and self-tuning quota field are then combined to form the diagnostic quota.

[0007] Preferably, in S2, the semi-healthy lock flag is a status flag indicating that the high-speed lathe is in a semi-healthy operating state, used to restrict the self-tuning behavior and control parameter update behavior of the high-speed lathe; the data structure of the semi-healthy lock flag includes a lock status field and a lock count field; the generation rule of the semi-healthy lock flag is as follows: when the diagnosable quota falls into a preset quota range defined by a first quota threshold and a second quota threshold and the number of consecutive statistical cycles corresponding to the lock count field reaches the lock holding condition, the lock status field is set to the locked state; when the diagnosable quota recovers to a safe quota range higher than the third quota threshold and the number of consecutive statistical cycles reaches the lock release condition, the lock status field is set to the unlocked state and the lock count field is cleared to zero.

[0008] Preferably, in S2, the machine tool status identifier includes a diagnosable quota, a semi-health lock identifier, and a status level field; the machine tool status identifier is divided into a normal status identifier, a semi-health status identifier, and a fault warning status identifier, and is classified by a combination of the diagnosable quota size and the semi-health lock identifier.

[0009] Preferably, in step S3, the self-tuning trigger condition is a self-tuning allowance condition and a self-tuning prohibition condition obtained by judging the current operating status of the high-speed lathe based on the diagnostic quota and the semi-healthy lock flag. This condition is used to determine whether the high-speed lathe performs normal self-tuning operation or enters self-tuning sandbox mode when a self-tuning request is triggered. The self-tuning trigger condition compares the diagnostic quota with the preset self-tuning allowance threshold and self-tuning prohibition threshold and performs logical judgment in combination with the lock status field corresponding to the semi-healthy lock flag. When the diagnostic quota is not lower than the self-tuning allowance threshold and the lock status field indicates an unlocked state, a normal self-tuning allowance flag is output. When the diagnostic quota is lower than the self-tuning prohibition threshold or the lock status field indicates a locked state, a normal self-tuning prohibition flag is output and the self-tuning sandbox mode is triggered.

[0010] Preferably, in S3, the self-tuning sandbox mode is a working mode in which a self-tuning algorithm is executed within a limited operating range to generate a set of candidate control parameters. It is used to evaluate the impact of control parameter adjustments on the operating behavior of high-speed lathes without affecting normal machining tasks and without directly writing the formal control parameters of the control system. The self-tuning sandbox mode defines an operating range consisting of a spindle speed range, a feed speed range, or load conditions, and executes a preset action sequence to drive the spindle and feed axis within this operating range. It collects observation data sequences from the production observation channel and the diagnostic observation channel within the limited operating range, inputs the observation data sequences into the self-tuning algorithm to calculate multiple candidate control parameters, associates and stores the candidate control parameters with the corresponding observation data to form a set of candidate control parameters, and keeps the formal control parameters of the control system unchanged throughout the execution of the self-tuning sandbox mode.

[0011] Preferably, in step S3, the high-speed lathe control parameters include spindle servo control parameters, feed servo control parameters, and compensation control parameters, used to determine the control law and compensation quantity of the high-speed lathe in each operating state, and also used to determine the final control system parameters of the high-speed lathe. The specific method for determining the final control system parameters of the high-speed lathe is as follows: the high-speed lathe control parameters updated by the self-tuning sandbox mode are used as candidate control parameters. Based on the observation data sequence of the diagnostic observation channel, the deviation data and control quantity changes under the action of the candidate control parameters are calculated and compared with the deviation data and control quantity changes under the action of the current control parameters. When the deviation data under the action of the candidate control parameters meets the preset diagnostic consistency judgment condition and the diagnosability quota is not lower than the quota lower limit threshold, the candidate control parameter is confirmed as the target control parameter and written into the formal control parameter storage area of ​​the control system. When the candidate control parameter does not meet the diagnostic consistency judgment condition, or causes the diagnosability quota to be lower than the quota lower limit threshold, the current control parameter is kept unchanged and the candidate control parameter is discarded.

[0012] Preferably, in step S4, the control and handling operation is a set of control actions that adjust the operating mode and control parameter boundaries of the high-speed lathe based on the machine tool status identifier and fault diagnosis information. This is used to limit the operating range of the high-speed lathe under different machine tool states and output corresponding control commands. Specifically, the high-speed lathe performs the control and handling operation as follows: when the machine tool status identifier is a normal status identifier, the control system maintains the current formal control parameters and executes the machining task in normal operating mode; when the machine tool status identifier is a semi-healthy status identifier, the control system applies derating restrictions to the spindle speed range and feed rate range and executes the machining task in restricted operating mode; when the machine tool status identifier is a fault warning status identifier, the control system outputs a stop control command based on the fault diagnosis information and prohibits further execution of the machining task.

[0013] On the other hand, the present invention provides an intelligent fault diagnosis system for high-speed lathes, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the aforementioned intelligent fault diagnosis method for high-speed lathes.

[0014] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects: 1. In this invention, based on the separation structure of the production observation channel and the diagnostic observation channel, as well as the computer mechanism of diagnostic quota and semi-health lock mark, it is possible to identify the semi-healthy state that has long relied on compensation control and self-tuning adjustment to maintain the high-speed lathe while still meeting the machining accuracy requirements. The information margin that can be used for fault diagnosis under the current control state is characterized in the form of quota, so as to avoid the compensation and self-tuning behavior completely canceling out the potential fault signs and causing the fault diagnosis results to be distorted. 2. In this invention, by introducing a self-tuning sandbox mode and combining it with the normal operation, derating operation and shutdown protection control operations driven by the machine tool status identifier, the self-tuning parameter adjustment is first verified for safety within a limited operating range before deciding whether to write it into the formal control parameters. While ensuring the availability of the self-tuning function, it controls the degree of erosion of diagnostic information and the operational risks, making it easier for high-speed lathes to maintain reasonable control performance and reduce the risk of sudden failure shutdowns during long-term operation. Attached Figure Description

[0015] Figure 1 This is a flowchart of one embodiment of the present invention. Detailed Implementation

[0016] Example 1, as Figure 1 As shown, the specific implementation steps of the intelligent fault diagnosis method for high-speed lathes proposed in this invention are as follows: S1. Collect the operating data of the high-speed lathe and construct the production observation channel and the diagnostic observation channel; apply compensation control, filtering control and self-tuning control to the production observation channel, and shield compensation control and self-tuning control to the diagnostic observation channel; S2. Real-time acquisition of observation data sequences from production observation channels and diagnostic observation channels; calculation of deviation data, compensation control quantity, and self-tuning adjustment quantity between the two channels; calculation of diagnostic quota based on deviation data, compensation control quantity, and self-tuning adjustment quantity and generation of semi-health lock-in flag as a new machine tool status flag. Wherein, the diagnostic quota is the quota data of the remaining degree of diagnostic information of the high-speed lathe under the current control state; the semi-health lock flag is a status flag indicating that the high-speed lathe is in a state without alarm but in a degraded state and restricts control behavior; the machine tool status flag includes the diagnostic quota and the semi-health lock flag. S3. Read the diagnostic quota and semi-health lock flag, and determine whether the high-speed lathe meets the self-tuning conditions based on the diagnostic quota and semi-health lock flag; perform normal self-tuning operation when the high-speed lathe meets the self-tuning conditions, and enter the self-tuning sandbox mode and update the high-speed lathe control parameters when the high-speed lathe does not meet the self-tuning conditions. The self-tuning sandbox mode is a self-tuning mode that performs self-tuning calculations within a limited operating range isolated from normal processing tasks, and does not directly modify the formal control parameters of the control system. S4. Based on the machine tool status identifier and the observation data sequence of the diagnostic observation channel, generate fault diagnosis information, and combine it with the updated high-speed lathe control parameters to perform control and handling operations on the high-speed lathe.

[0017] In this embodiment S1, the production observation channel and the diagnostic observation channel are two observation paths formed by splitting the high-speed lathe operation data based on a unified sampling time reference. The production observation channel is used to output the production observation data sequence under the conditions of performing compensation control, filtering control and self-tuning control, and the diagnostic observation channel is used to output the diagnostic observation data sequence under the conditions of performing signal preprocessing operation. The production observation channel and the diagnostic observation channel output the observation data sequence with the unified sampling time reference as the index.

[0018] In this embodiment S1, the high-speed lathe operation data is uniformly collected from the spindle drive, feed servo drive, and position feedback unit through the data acquisition task of the CNC system. The operation data includes speed sampling data reflecting the spindle speed state, position sampling data and speed sampling data reflecting the feed axis motion state, follower error sampling data reflecting the feed error, current sampling data reflecting the spindle and feed load state, and encoder or scale sampling data reflecting the position feedback. The CNC system completes the sampling of the above operation data within a fixed control cycle, and uses the sampling time within the control cycle as a unified sampling time reference, assigning a unique time index field to each set of operation data. Under the unified sampling time reference, the control system constructs the operation data collected in each control cycle into a sampling data sequence arranged in ascending order of time index in the internal memory, providing a unified data source and a unified time reference for the construction of production observation channels and diagnostic observation channels.

[0019] In this embodiment S1, the production observation channel and the diagnostic observation channel are formed within the software logic layer of the CNC system by splitting the same set of sampled data. After the control system completes the acquisition of the running data, it copies a sampled data sequence under a unified sampling time reference, using one copy as the input of the production observation channel and the other copy as the input of the diagnostic observation channel. The production observation channel processes the input sampled data sequence in the order of signal preprocessing, compensation control, filtering control, and self-tuning control. Signal preprocessing is used to complete noise reduction filtering and range normalization. Compensation control is used to superimpose thermal error compensation, pitch compensation, and other compensation amounts on the sampled data. Filtering control is used to perform vibration suppression filtering operations as needed. Self-tuning control is used to adjust the control parameters according to the current operating state to generate new control commands. After completing the above processing, the production observation channel outputs the production observation data sequence and uses the production observation data sequence as the basis for generating spindle control commands and feed axis control commands, thereby completing the conversion from the original sampled data to the observation data used for machining process control without changing the unified sampling time reference.

[0020] In this embodiment S1, after receiving the sampled data sequence under the same unified sampling time reference as the production observation channel, the diagnostic observation channel only performs the same signal preprocessing operations as the production observation channel, which are used to perform noise filtering, range normalization, and necessary coordinate transformation on the sampled data. It does not perform compensation control or self-tuning control in the diagnostic observation channel, nor does it apply the compensation amount or self-tuning result to the sampled data in the diagnostic observation channel. After completing the signal preprocessing, the diagnostic observation channel outputs the diagnostic observation data sequence. The control system, using the unified sampling time reference as an index, simultaneously records the production observation corresponding to that control cycle within each control cycle. Both production observation data and diagnostic observation data are constructed as paired observation records with time index as the key and production observation data and diagnostic observation data as the value. They are cached using a sequential storage structure or a circular buffer structure to ensure that production observation data and diagnostic observation data under any time index correspond one-to-one and have strict sampling timing consistency. Spindle control commands and feed axis control commands are generated only by the control calculation link corresponding to the production observation channel and act on the actuator. The diagnostic observation channel only serves as the output path for diagnostic data and does not participate in the formation of control commands. This ensures that the control behavior and diagnostic observation are logically independent but synchronized in sampling time.

[0021] In this embodiment S2, the diagnostic quota is quota data calculated based on the deviation data, compensation control quantity, and self-tuning adjustment quantity of the production observation channel and the diagnostic observation channel within a preset statistical time window. It is used to characterize the degree of remaining diagnostic information of the high-speed lathe under the current control state. The data structure of the diagnostic quota includes a deviation quota field, a compensation quota field, and a self-tuning quota field. The diagnostic quota is constructed as follows: the deviation data, compensation control quantity, and self-tuning adjustment quantity are calculated separately in each statistical time window, and the deviation quota field, compensation quota field, and self-tuning quota field are obtained by combining them according to preset weight coefficients. The deviation quota field, compensation quota field, and self-tuning quota field are then combined to form the diagnostic quota.

[0022] In this embodiment S2, the control system reads the observation data sequences of the production observation channel and the diagnostic observation channel in real time under a unified sampling time reference. Within each control cycle, it calculates the deviation data based on the production observation data and diagnostic observation data at the same time index position. The deviation data can be the difference of a single observation or a difference vector composed of the differences of the components of a multi-dimensional observation vector. Simultaneously, it extracts compensation control quantities and self-tuning adjustment quantities from the control calculation chain corresponding to the production observation channel. The compensation control quantity reflects the thermal error compensation, pitch compensation, or other geometric error compensation quantities superimposed on the control command in the current control cycle. The self-tuning adjustment quantity is used to reflect... The adjustment magnitude or direction of the self-tuning algorithm on the control parameters within the current control cycle is reflected. The control system constructs a statistical time window of fixed length on the time axis, and organizes the deviation data, compensation control quantity and self-tuning adjustment quantity corresponding to multiple consecutive control cycles into a sequence of deviation data, a sequence of compensation control quantity and a sequence of self-tuning adjustment quantity within the window in chronological order. At the end of each statistical time window, at least one statistical quantity is calculated for the deviation data, compensation control quantity and self-tuning adjustment quantity within the window. The statistical quantity may include one or more of the following: mean, absolute average, variance or extreme value, used to reflect the overall change level of deviation data and control behavior within the window.

[0023] In this embodiment S2, the diagnosable quota is represented by a data structure including a deviation quota field, a compensation quota field, and a self-tuning quota field. The deviation quota field records the normalized result of the statistical quantity corresponding to the deviation data within the statistical time window. The compensation quota field records the normalized result of the statistical quantity corresponding to the compensation control quantity within the statistical time window. The self-tuning quota field records the normalized result of the statistical quantity corresponding to the self-tuning adjustment quantity within the statistical time window. The control system pre-configures the weight coefficients of the deviation quota field, the compensation quota field, and the self-tuning quota field according to the machine tool type and operating condition characteristics. At the end of each statistical time window, the deviation data statistical quantity is multiplied by the corresponding deviation weight coefficient to generate the deviation quota field value, and the compensation control quantity statistical quantity is multiplied by the corresponding compensation weight coefficient to generate the compensation quota field value. The self-tuning adjustment quantity statistics are multiplied by the corresponding self-tuning weight coefficient to generate the self-tuning quota field value. Then, the three fields are combined according to the preset combination rules to form the diagnostic quota. The numerical range of the diagnostic quota is limited to a fixed interval through normalization processing, which is used to uniformly represent the degree of remaining diagnostic information under the current control state. When the deviation quota field value is high and the compensation quota field and self-tuning quota field values ​​are within the normal range, the diagnostic quota is close to the full quota value. When the deviation quota field value is low and the compensation quota field and self-tuning quota field values ​​are high, the diagnostic quota is close to the low quota value. After completing the quota calculation, the control system updates and stores the diagnostic quota together with the machine tool status identifier recorded in the previous statistical time window to form the machine tool status identifier record for the current period.

[0024] In this embodiment S2, the semi-healthy lock flag is a status flag indicating that the high-speed lathe is in a semi-healthy operating state, used to restrict the self-tuning behavior and control parameter update behavior of the high-speed lathe; the data structure of the semi-healthy lock flag includes a lock status field and a lock count field; the generation rule of the semi-healthy lock flag is as follows: when the diagnosable quota falls into a preset quota range defined by the first quota threshold and the second quota threshold and the number of consecutive statistical cycles corresponding to the lock count field reaches the lock holding condition, the lock status field is set to the locked state; when the diagnosable quota recovers to a safe quota range higher than the third quota threshold and the number of consecutive statistical cycles reaches the lock release condition, the lock status field is set to the unlocked state and the lock count field is cleared to zero.

[0025] In this embodiment S2, the semi-healthy lock identifier is represented by a data structure including a lock status field and a lock count field. The lock status field indicates whether the machine tool is currently in a semi-healthy lock state, and the lock count field records the number of consecutive statistical time windows that meet the lock conditions. The control system sets a first quota threshold and a second quota threshold for the diagnosable quota to limit a preset quota range, and sets a third quota threshold to limit the lower limit of the safe quota range. When the diagnosable quota value corresponding to the current statistical time window falls into the preset quota range limited by the first quota threshold and the second quota threshold, the lock count field is incremented by one. When the number of consecutive statistical time windows accumulated in the lock count field reaches the preset lock holding period, the lock is released. When the status field is set to locked, it indicates that the high-speed lathe has entered a semi-healthy locked state. When the diagnostic quota value corresponding to the subsequent statistical time window is continuously higher than the third quota threshold and maintains the preset unlocking period, the lock status field is set to unlocked and the lock count field is cleared to zero, indicating that the high-speed lathe has exited the semi-healthy locked state. When the lock status field is locked, the control system applies behavioral constraints to the self-tuning trigger request and the control parameter update request. The self-tuning request is only allowed to enter the restricted execution mode, and the control parameter update request is only executed when additional check conditions are met. Thus, the semi-healthy lock flag plays a role in constraining and controlling the self-tuning behavior and the control parameter update behavior in the machine tool status flag.

[0026] In this embodiment S2, the machine tool status identifier includes a diagnosable quota, a semi-health lock identifier, and a status level field; the machine tool status identifier is divided into a normal status identifier, a semi-health status identifier, and a fault warning status identifier, and is classified by a combination of the diagnosable quota size and the semi-health lock identifier.

[0027] In this embodiment S2, the status level field is used to indicate whether the high-speed lathe is in a normal state, a semi-healthy state, or a fault warning state. The machine tool status identifier is classified by a combination of the diagnostic quota size and the semi-healthy lock identifier. The specific method is as follows: when the diagnostic quota is higher than the first status quota threshold and the lock status field corresponding to the semi-healthy lock identifier is unlocked, the status level field indicates a normal state identifier. When the diagnostic quota is between the first status quota threshold and the second status quota threshold, or when the lock status field corresponding to the semi-healthy lock identifier is locked, the status level field indicates a semi-healthy state identifier. When the diagnostic quota is lower than the second status quota threshold, the status level field indicates a fault warning state identifier.

[0028] In this embodiment S3, the self-tuning trigger condition is a self-tuning allowable condition and a self-tuning prohibitive condition obtained by judging the current operating status of the high-speed lathe based on the diagnostic quota and the semi-healthy lock flag. It is used to determine whether the high-speed lathe performs normal self-tuning operation or enters self-tuning sandbox mode when a self-tuning request is triggered. The self-tuning trigger condition compares the diagnostic quota with the preset self-tuning allowable threshold and self-tuning prohibitive threshold and performs logical judgment in combination with the lock status field corresponding to the semi-healthy lock flag. When the diagnostic quota is not lower than the self-tuning allowable threshold and the lock status field indicates an unlocked state, a normal self-tuning allowable flag is output. When the diagnostic quota is lower than the self-tuning prohibitive threshold or the lock status field indicates a locked state, a normal self-tuning prohibitive flag is output and the self-tuning sandbox mode is triggered.

[0029] In this embodiment S3, the self-tuning request is initiated by the self-tuning scheduling strategy within the CNC system or by the manual operation interface. When a self-tuning request is detected, the control system reads the diagnostic quota corresponding to the current statistical time window and the semi-healthy lock flag in the machine tool status identifier based on a unified sampling time reference. It obtains the lock status field contained in the semi-healthy lock flag and simultaneously retrieves the pre-set self-tuning allowable threshold and self-tuning prohibition threshold. By comparing the diagnostic quota with the self-tuning allowable threshold and self-tuning prohibition threshold and combining it with the lock status field for logical judgment, a self-tuning trigger condition is established. Among them, the self-tuning trigger condition is established when the diagnostic quota is not lower than the self-tuning allowable threshold. When the value and the lock status field indicates an unlocked state, the self-tuning trigger condition gives a normal self-tuning allowed result. When the diagnostic quota is lower than the self-tuning prohibited threshold or the lock status field indicates a locked state, the self-tuning trigger condition gives a normal self-tuning prohibited result and specifies the use of self-tuning sandbox mode. When the diagnostic quota is between the self-tuning allowed threshold and the self-tuning prohibited threshold, the self-tuning trigger condition classifies the self-tuning request into a restricted execution state according to the status level field in the machine tool status identifier and the priority rules preset for the current machining task level, so that when the remaining diagnostic information is in the middle range, the selection of the self-tuning mode is controlled by strategic decision-making.

[0030] In this embodiment S3, the self-tuning allowable threshold is used to limit the minimum remaining level of diagnostic information required for the diagnostic quota to reach normal self-tuning, and the self-tuning prohibition threshold is used to limit the lower limit level at which the diagnostic quota decreases to a level that cannot support normal self-tuning. The value of the self-tuning allowable threshold is higher than the self-tuning prohibition threshold, which is used to form a transition range for the self-tuning strategy between the two. When the diagnostic quota is within the transition range, the self-tuning trigger condition is determined based on the machine tool status and machining task level indicated by the status level field. When the status level field indicates a positive value... When the status is normal and the processing task level is low, the self-tuning trigger condition can temporarily classify the self-tuning request as a normal self-tuning allowed result. When the status level field indicates a semi-healthy status or the processing task level is high, the self-tuning trigger condition will classify the self-tuning request as a normal self-tuning prohibited result and prioritize the self-tuning sandbox mode. Thus, when the diagnostic quota is in the middle range, the self-tuning trigger strategy is controlled by introducing status information and task priority, avoiding frequent execution of normal self-tuning calculations when diagnostic information is insufficient or the task level is high.

[0031] In this embodiment S3, the self-tuning sandbox mode is a working mode that executes a self-tuning algorithm within a limited operating range to generate a set of candidate control parameters. It is used to evaluate the impact of control parameter adjustments on the operating behavior of high-speed lathes without affecting normal machining tasks and without directly writing the formal control parameters of the control system. The self-tuning sandbox mode limits the operating range to a range consisting of spindle speed range, feed speed range, or load conditions, and executes a preset action sequence to drive the spindle and feed axis within this operating range. It collects the observation data sequence of the production observation channel and the diagnostic observation channel within the limited operating range, inputs the observation data sequence into the self-tuning algorithm to calculate multiple candidate control parameters, associates and stores the candidate control parameters with the corresponding observation data to form a set of candidate control parameters, and keeps the formal control parameters of the control system unchanged throughout the execution of the self-tuning sandbox mode.

[0032] In this embodiment S3, the self-tuning sandbox mode is a working mode that executes the self-tuning algorithm under limited operating conditions. It is used to generate candidate control parameters and evaluate the impact of control parameter adjustments on the high-speed lathe's operating behavior while keeping the formal control parameters of the control system unchanged. The self-tuning sandbox mode determines the limited operating range by pre-configuring a sandbox operating parameter set in the control system. The limited operating range includes the spindle speed range, feed rate range, and load condition range of the spindle and feed axes, preferentially selecting the combination of speed and feed rate under no-load or light-load conditions. After the self-tuning sandbox mode is activated, the control system issues a preset action sequence within the limited operating range. The preset action sequence includes a set of standard parameters for stimulating the spindle servo response and feed servo response. The system uses a set of motion instructions to drive the spindle and feed axis to perform the operation process according to preset actions. During the execution of the preset actions, observation data sequences are continuously collected through the production observation channel and the diagnostic observation channel. The observation data sequences collected within the limited operating range are provided to the self-tuning algorithm. The self-tuning algorithm analyzes the observation data sequences while keeping the formal control parameters unchanged, and calculates multiple sets of candidate control parameters. Each set of candidate control parameters and its corresponding observation data statistical characteristics are recorded together to construct a set of candidate control parameters. During the entire process of running in the self-tuning sandbox mode, the control commands issued by the control system to the actuators are still generated based on the current formal control parameters to ensure that the self-tuning sandbox mode does not change the control parameters used in normal machining tasks.

[0033] In this embodiment S3, the sandbox operation parameter set is set through the parameter configuration interface of the CNC system. The sandbox operation parameter set includes parameter fields such as the lower limit of spindle speed, the upper limit of spindle speed, the lower limit of feed rate, the upper limit of feed rate, the upper limit of spindle load, the upper limit of feed axis load, and the allowable duration of sandbox operation. The lower limit and upper limit of spindle speed are used to limit the spindle speed range in sandbox mode, so that the spindle operates within a safe range below the normal machining speed. The lower limit and upper limit of feed rate are used to limit the feed rate range of each feed axis, so that the feed motion is in a low-speed or idle state that does not affect the workpiece quality. The load limit and feed axis load limit are used to limit the maximum allowable load level in sandbox mode. The allowable duration of sandbox operation is used to limit the maximum duration of continuous operation in a single sandbox mode. Before entering the self-tuning sandbox mode, the control system reads the sandbox operation parameter set and verifies the current machine tool operating status. The self-tuning sandbox mode is only started when the current spindle speed, feed rate and load level can be adjusted to the range limited by the sandbox operation parameter set and the remaining available time is greater than the allowable duration of sandbox operation. Otherwise, the sandbox operation request is postponed or rejected, thereby avoiding the execution of sandbox self-tuning calculation under unsuitable working conditions.

[0034] In this embodiment S3, the preset action sequence is set by the control system during the machine tool parameter configuration stage. The preset action sequence includes a spindle acceleration / deceleration action subsequence and a feed axis reciprocating motion subsequence. The spindle acceleration / deceleration action subsequence defines multiple target speed levels within the spindle speed range defined by the sandbox operation parameter set. The control system controls the spindle to accelerate sequentially from low speed to each target speed level according to a fixed time interval or a fixed speed gradient, and then maintains a constant speed for a short time before decelerating back to the initial speed. The feed axis reciprocating motion subsequence defines multiple target position intervals and target feed speeds within the feed speed range defined by the sandbox operation parameter set. The control system controls one or more... Each feed axis performs reciprocating motion between these target position intervals, forming representative servo response scenarios by combining different feed speed configurations. During the execution of the preset action sequence, the control system records operating data such as spindle speed, spindle current, feed axis position, feed axis speed, feed axis current, and compensation amount in each control cycle, and synchronously outputs the observation data sequence through the production observation channel and the diagnostic observation channel. The self-tuning algorithm divides the sampling interval according to the execution stage of the preset action sequence, and analyzes the observation data sequence corresponding to each sampling interval as response data under different operating conditions, thereby improving the coverage of candidate control parameters to actual operating conditions.

[0035] In this embodiment S3, the candidate control parameter set is managed using a record table structure. Each candidate control parameter record includes a candidate spindle servo control parameter, a candidate feed servo control parameter, a candidate compensation control parameter, a time index of the generation time, a corresponding preset action sequence stage identifier, and statistical feature fields of the observation data for that stage. The candidate spindle servo control parameter describes the combination of parameters such as gain and time constant of the spindle speed loop and position loop. The candidate feed servo control parameter describes the combination of parameters such as gain and feedforward coefficient of each feed axis speed loop and position loop. The candidate compensation control parameter describes the coefficients or corrections for thermal error compensation and pitch compensation. The time index is used to mark the timing of candidate control parameter generation, and the stage identifier is used to mark the candidate control parameter as... In which action phase of the preset action sequence is the candidate control parameter generated? The observation data statistical feature field is used to record the deviation data statistics and control quantity statistics corresponding to the generation phase of the candidate control parameter. During the execution of the self-tuning sandbox mode, the control system writes each set of candidate control parameters output by the self-tuning algorithm, along with its corresponding time index, stage identifier, and observation data statistical features, into the candidate control parameter set. The candidate control parameters obtained from the same round of self-tuning sandbox mode or multiple rounds of self-tuning sandbox mode are uniformly managed through the candidate control parameter set. After the sandbox mode ends, the screening, evaluation, and target control parameter determination processes are performed based on the candidate control parameter set. At the same time, the formal control parameters of the control system remain unchanged throughout the entire candidate control parameter generation and management process.

[0036] In this embodiment S3, the high-speed lathe control parameters include spindle servo control parameters, feed servo control parameters, and compensation control parameters. These parameters are used to determine the control law and compensation amount of the high-speed lathe under various operating states, and also to determine the final control system parameters of the high-speed lathe. The specific method for determining the final control system parameters of the high-speed lathe is as follows: the high-speed lathe control parameters updated by the self-tuning sandbox mode are used as candidate control parameters. Based on the observation data sequence of the diagnostic observation channel, the deviation data and control quantity changes under the action of the candidate control parameters are calculated and compared with the deviation data and control quantity changes under the action of the current control parameters. When the deviation data under the action of the candidate control parameters meets the preset diagnostic consistency judgment condition and the diagnosability quota is not lower than the quota lower limit threshold, the candidate control parameter is confirmed as the target control parameter and written into the formal control parameter storage area of ​​the control system. When the candidate control parameter does not meet the diagnostic consistency judgment condition, or causes the diagnosability quota to be lower than the quota lower limit threshold, the current control parameter remains unchanged and the candidate control parameter is discarded.

[0037] In this embodiment S3, the high-speed lathe control parameters include spindle servo control parameters, feed servo control parameters, and compensation control parameters. The spindle servo control parameters are used to determine the control laws of the spindle speed loop and position loop. The feed servo control parameters are used to determine the control laws of the speed loop and position loop of each feed axis. The compensation control parameters are used to determine the calculation method and amplitude of compensation quantities such as thermal error compensation and pitch compensation. After the self-tuning sandbox mode ends, the control system reads candidate control parameters one by one from the candidate control parameter set and temporarily loads the candidate control parameters into the diagnostic evaluation environment. The diagnostic evaluation environment calculates the deviation data and control quantity changes under the action of the candidate control parameters based on the observation data sequence of the diagnostic observation channel, and compares and analyzes them with the deviation data and control quantity changes under the action of the current formal control parameters. By setting diagnostic consistency judgment conditions, it is determined whether the candidate control parameters maintain or improve the deviation characteristics while keeping the control quantity changes within the allowable range. The diagnostic consistency judgment conditions may include constraints such as the difference between the deviation data statistics and the corresponding deviation data statistics of the current formal control parameters not exceeding a preset deviation threshold and the control quantity changes not exceeding a preset change threshold.

[0038] In this embodiment S3, the control system configures a quota lower limit threshold for the candidate control parameter screening process. The quota lower limit threshold is used to limit the lower bound of the allowable reduction of the diagnostic quota when using candidate control parameters. While performing diagnostic consistency judgment on candidate control parameters, the control system recalculates the diagnostic quota based on the observation data under the action of candidate control parameters. When the deviation data under the action of candidate control parameters meets the diagnostic consistency judgment condition and the corresponding diagnostic quota is not lower than the quota lower limit threshold, the candidate control parameter is confirmed as the target control parameter and written into the formal control parameter storage area of ​​the control system to replace the original formal control parameter. When the candidate control parameter does not meet the diagnostic consistency judgment condition or causes the diagnostic quota to be lower than the quota lower limit threshold, the control system keeps the current formal control parameter unchanged, marks the candidate control parameter as invalid in the candidate control parameter set and no longer participates in subsequent screening. This ensures that the update of the high-speed lathe control parameters meets the control performance requirements while maintaining the remaining degree of diagnostic information within the allowable range and avoiding the frequent use of parameter combinations that are detrimental to diagnosis.

[0039] In this embodiment S4, the control and handling operation is a set of control actions that adjust the operating mode and control parameter boundaries of the high-speed lathe based on the machine tool status identifier and fault diagnosis information. This is used to limit the operating range of the high-speed lathe under different machine tool states and output corresponding control commands. Specifically, the high-speed lathe performs the control and handling operation as follows: When the machine tool status identifier is a normal status identifier, the control system maintains the current formal control parameters and executes the machining task in normal operating mode; when the machine tool status identifier is a semi-healthy status identifier, the control system applies derating restrictions to the spindle speed range and feed rate range and executes the machining task in restricted operating mode; when the machine tool status identifier is a fault warning status identifier, the control system outputs a stop control command based on the fault diagnosis information and prohibits further execution of the machining task.

[0040] In this embodiment S4, fault diagnosis information is generated based on the observation data sequence of the diagnostic observation channel and the current machine tool status identifier. The fault diagnosis information includes a fault type identifier, a fault severity level, an associated part identifier, a time index field, a snapshot of the corresponding machine tool status identifier, and a diagnostic feature summary field. The fault type identifier is used to indicate different types of fault categories such as deviation abnormality, vibration abnormality, servo following abnormality, or compensation abnormality. The fault severity level is used to distinguish different severity levels such as warning level, restriction level, and shutdown level. The associated part identifier is used to identify the spindle system, a certain feed axis, or a combination component. The time index field is used to mark the time when the fault diagnosis information is generated. The snapshot of the corresponding machine tool status identifier is used to record the machine tool status identifier content when the fault diagnosis information is generated. The diagnostic feature summary field is used to record information such as deviation statistics, vibration characteristics, and compensation behavior characteristics extracted from the observation data sequence of the diagnostic observation channel. The control system comprehensively analyzes the long-term trend of deviation data, the change of diagnosable quota, and the status of semi-health lock identifier in the diagnostic observation channel, and maps the observation features that meet the preset abnormality judgment conditions to the corresponding fault type identifier and fault severity level, thereby constructing a structured fault diagnosis information record.

[0041] In embodiment S4, after generating fault diagnosis information, the control system jointly judges the fault diagnosis information and the machine tool status identifier to determine the strategy category and specific action content of the control and handling operation. The machine tool status identifier provides the current operating status level information, including the normal status identifier, the semi-healthy status identifier, and the fault warning status identifier. The fault diagnosis information provides the specific fault type, severity, and associated parts. The control system first determines whether the control and handling operation belongs to the normal operation strategy, the derating operation strategy, or the shutdown protection strategy based on the machine tool status identifier. Then, it determines the scope of the control and handling operation and the direction of the control quantity adjustment based on the fault type identifier and the associated part identifier in the fault diagnosis information. For example, when the semi-healthy state and the fault type identifier indicates a deviation abnormality related to the spindle, the upper limit of the spindle speed is limited first. When the semi-healthy state and the fault type identifier indicates a servo follow abnormality of a certain feed axis, the upper limit of the feed speed of the corresponding feed axis is limited first. Through the above joint judgment mechanism, the control and handling operation is kept consistent in terms of both status level and fault type.

[0042] In this embodiment S4, when the machine tool status identifier is a normal status identifier, the control system sets the control handling operation to a record-only handling, maintaining the current formal control parameters unchanged, keeping the upper limit of spindle speed, upper limit of feed speed, and compensation control parameters unchanged, allowing the execution of all issued machining programs and newly issued machining programs, and only writing fault diagnosis information into the status record area. The status record area assigns a record index field to each fault diagnosis information and associates it with the corresponding time index and machine tool status identifier snapshot, which is used to retain the diagnosis results and operating status trajectory without changing the current operating mode. Under the normal status identifier, the operation interface can display a fault diagnosis information summary and status record index, prompting the operator that the current operating state is well diagnosable and does not require intervention.

[0043] In embodiment S4, when the machine tool status is marked as semi-healthy, the control system sets the control operation to derated operation. Based on the fault severity level and associated component identifier in the fault diagnosis information, the operating limit parameters of the high-speed lathe are adjusted. For cases where the fault severity is at the limit level and the associated component is the spindle system, the current upper limit of the spindle speed is reduced by a preset proportional factor, limiting the usable spindle speed range to the derated operating range. For cases where the associated component is a feed axis, the upper limit of the feed rate of the corresponding feed axis is reduced by a preset proportional factor. Simultaneously, the machining task scheduling queue is scanned, and machining tasks with high load characteristics or high cutting parameter tags are marked as prohibited from being issued. Only currently running machining tasks and machining tasks marked as low load are allowed to be executed. Furthermore, the control system generates a maintenance prompt record, which includes the suggested maintenance operation type, suggested execution time, associated fault type identifier, and associated component identifier. The maintenance prompt record is associated with the current machining task identifier and stored so that maintenance personnel can be prompted to perform corresponding inspection and maintenance operations after the machining task is completed.

[0044] In embodiment S4, when the machine tool status is marked as a fault warning status, the control system sets the control operation to a shutdown protection type. Following the preset shutdown procedure, it sequentially executes load reduction and shutdown actions. First, it sends a stop interpolation command to the CNC system to stop the interpolation movement of each feed axis. If necessary, it performs a safe tool retraction action to move the tool from the workpiece surface to a safe position. Second, it sends a spindle deceleration and shutdown command to control the spindle to decelerate from its current speed to zero speed and enter a braking state. Then, it cancels the current machining cycle, marks the currently executed machining program status as a stopped state, and records the reason for the stop. Finally, it blocks new start commands by setting a run permission flag, making the start buttons such as automatic run and single-segment run on the operation interface unavailable until the fault status is confirmed and reset. While executing the above shutdown procedure, the control system writes the fault diagnosis information, shutdown event identifier, current formal control parameter snapshot, current machine tool status identifier snapshot, and shutdown time index into the event record area. The event record area stores all shutdown event information to support subsequent analysis and tracing of the fault evolution process and control strategy.

[0045] Example 2: The present invention proposes a high-speed lathe intelligent fault diagnosis system, which is applied to the high-speed lathe intelligent fault diagnosis method proposed in Example 1. It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the high-speed lathe intelligent fault diagnosis method in Example 1.

[0046] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A method for intelligent fault diagnosis of a high-speed lathe, characterized in that, Includes the following steps: S1. Collect operating data of high-speed lathes to construct production observation channels and diagnostic observation channels; The production observation channel is subjected to load compensation control, filtering control and self-tuning control; the diagnostic observation channel is subjected to shielding compensation control and self-tuning control. S2. Real-time acquisition of observation data sequences from production observation channels and diagnostic observation channels; calculation of deviation data, compensation control quantity, and self-tuning adjustment quantity between the two channels; calculation of diagnostic quota based on deviation data, compensation control quantity, and self-tuning adjustment quantity and generation of semi-health lock-in flag as a new machine tool status flag. Wherein, the diagnostic quota is the quota data of the remaining degree of diagnostic information of the high-speed lathe under the current control state; the semi-health lock flag is a status flag indicating that the high-speed lathe is in a state without alarm but in a degraded state and restricts control behavior; the machine tool status flag includes the diagnostic quota and the semi-health lock flag. In S2, the semi-healthy lockout flag is a status flag indicating that the high-speed lathe is in a semi-healthy operating state, used to limit the self-tuning behavior and control parameter update behavior of the high-speed lathe; the data structure of the semi-healthy lockout flag includes a lockout status field and a lockout count field; the generation rule of the semi-healthy lockout flag is as follows: when the diagnosable quota falls into the preset quota range defined by the first quota threshold and the second quota threshold and the number of consecutive statistical cycles corresponding to the lockout count field reaches the lockout holding condition, the lockout status field is set to the locked state; when the diagnosable quota recovers to the safe quota range higher than the third quota threshold and the number of consecutive statistical cycles reaches the lockout release condition, the lockout status field is set to the unlocked state and the lockout count field is cleared to zero. S3. Read the diagnostic quota and semi-health lock flag, and determine whether the high-speed lathe meets the self-tuning conditions based on the diagnostic quota and semi-health lock flag; perform normal self-tuning operation when the high-speed lathe meets the self-tuning conditions, and enter the self-tuning sandbox mode and update the high-speed lathe control parameters when the high-speed lathe does not meet the self-tuning conditions. The self-tuning sandbox mode is a self-tuning mode that performs self-tuning calculations within a limited operating range isolated from normal processing tasks, and does not directly modify the formal control parameters of the control system. S4. Based on the machine tool status identifier and the observation data sequence of the diagnostic observation channel, generate fault diagnosis information, and combine it with the updated high-speed lathe control parameters to perform control and handling operations on the high-speed lathe.

2. The intelligent fault diagnosis method for high-speed lathes according to claim 1, characterized in that: In S1, the production observation channel and the diagnostic observation channel are two observation paths formed by splitting the high-speed lathe operation data based on a unified sampling time reference. The production observation channel is used to output the production observation data sequence under the conditions of performing compensation control, filtering control and self-tuning control, and the diagnostic observation channel is used to output the diagnostic observation data sequence under the conditions of performing signal preprocessing operation. The production observation channel and the diagnostic observation channel output the observation data sequence with the unified sampling time reference as the index.

3. The intelligent fault diagnosis method for high-speed lathes according to claim 2, characterized in that: In S2, the diagnostic quota is quota data calculated based on the deviation data, compensation control quantity, and self-tuning adjustment quantity of the production observation channel and the diagnostic observation channel within a preset statistical time window. It is used to characterize the degree of remaining diagnostic information of the high-speed lathe under the current control state. The data structure of the diagnostic quota includes a deviation quota field, a compensation quota field, and a self-tuning quota field. The diagnostic quota is constructed as follows: in each statistical time window, the deviation data, compensation control quantity, and self-tuning adjustment quantity are calculated respectively, and the deviation quota field, compensation quota field, and self-tuning quota field are obtained by combining them according to preset weight coefficients. The deviation quota field, compensation quota field, and self-tuning quota field are then combined to form the diagnostic quota.

4. The intelligent fault diagnosis method for high-speed lathes according to claim 3, characterized in that: In S2, the machine tool status identifier includes a diagnosable quota, a semi-health lock identifier, and a status level field; the machine tool status identifier is divided into a normal status identifier, a semi-health status identifier, and a fault warning status identifier, and is classified by a combination of the diagnosable quota size and the semi-health lock identifier.

5. The intelligent fault diagnosis method for high-speed lathes according to claim 4, characterized in that: In S3, the self-tuning triggering conditions are self-tuning allowable conditions and self-tuning prohibitive conditions obtained by judging the current operating status of the high-speed lathe based on the diagnostic quota and semi-health lock flag. These conditions are used to determine whether the high-speed lathe performs normal self-tuning operation or enters self-tuning sandbox mode when the self-tuning request is triggered. The self-tuning trigger condition compares the diagnosable quota with the preset self-tuning allow threshold and self-tuning prohibit threshold, and performs logical judgment based on the lock status field corresponding to the semi-healthy lock flag. When the diagnosable quota is not lower than the self-tuning allow threshold and the lock status field indicates an unlocked state, a normal self-tuning allow flag is output. When the diagnosable quota is lower than the self-tuning prohibit threshold or the lock status field indicates a locked state, a normal self-tuning prohibit flag is output and the self-tuning sandbox mode is triggered.

6. The intelligent fault diagnosis method for high-speed lathes according to claim 5, characterized in that: In S3, the self-tuning sandbox mode is a working mode that executes a self-tuning algorithm within a limited operating range to generate a set of candidate control parameters. It is used to evaluate the impact of control parameter adjustments on the operating behavior of high-speed lathes without affecting normal machining tasks and without directly writing the formal control parameters of the control system. The self-tuning sandbox mode limits the operating range to a range consisting of a spindle speed range, a feed speed range, or load conditions, and executes a preset action sequence to drive the spindle and feed axis within this operating range. It collects the observation data sequence of the production observation channel and the diagnostic observation channel within the limited operating range, inputs the observation data sequence into the self-tuning algorithm to calculate multiple candidate control parameters, associates and stores the candidate control parameters with the corresponding observation data to form a set of candidate control parameters, and keeps the formal control parameters of the control system unchanged throughout the execution of the self-tuning sandbox mode.

7. The intelligent fault diagnosis method for high-speed lathes according to claim 6, characterized in that: In S3, the high-speed lathe control parameters include spindle servo control parameters, feed servo control parameters, and compensation control parameters. These parameters are used to determine the control law and compensation quantity of the high-speed lathe under various operating states, and also to determine the final control system parameters of the high-speed lathe. The specific method for determining the final control system parameters of the high-speed lathe is as follows: the high-speed lathe control parameters updated in the self-tuning sandbox mode are used as candidate control parameters. Based on the observation data sequence of the diagnostic observation channel, the deviation data and control quantity changes under the action of the candidate control parameters are calculated and compared with the deviation data and control quantity changes under the action of the current control parameters. When the deviation data under the action of the candidate control parameters meets the preset diagnostic consistency judgment condition and the diagnosability quota is not lower than the quota lower limit threshold, the candidate control parameter is confirmed as the target control parameter and written into the formal control parameter storage area of ​​the control system. When the candidate control parameter does not meet the diagnostic consistency judgment condition, or causes the diagnosability quota to be lower than the quota lower limit threshold, the current control parameter remains unchanged and the candidate control parameter is discarded.

8. The intelligent fault diagnosis method for high-speed lathes according to claim 7, characterized in that: In S4, the control and handling operation is a set of control actions that adjust the operating mode and control parameter boundaries of the high-speed lathe based on the machine tool status identifier and fault diagnosis information. This is used to limit the operating range of the high-speed lathe under different machine tool states and output corresponding control commands. Specifically, the high-speed lathe performs the control and handling operation as follows: When the machine tool status identifier is a normal status identifier, the control system maintains the current formal control parameters and executes the machining task in normal operating mode; when the machine tool status identifier is a semi-healthy status identifier, the control system applies derating restrictions to the spindle speed range and feed rate range and executes the machining task in restricted operating mode; when the machine tool status identifier is a fault warning status identifier, the control system outputs a stop control command based on the fault diagnosis information and prohibits further execution of the machining task.

9. A high-speed lathe intelligent fault diagnosis system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes a computer program to implement the intelligent fault diagnosis method for high-speed lathes as described in any one of claims 1-8.

Images