Intelligent diagnosis system for abnormal working state of spindle system of five-axis machining center

Abstract

The invention relates to the technical field of diagnosis control, and discloses an intelligent diagnosis system for working state abnormity of a five-axis machining center spindle system. Comprising a period calculation module, a feature acquisition module, an intelligent diagnosis module and a state adjustment module, and is characterized in that deadline change parameters of a period period are acquired, an expansion period is planned, comprehensive state features of the expansion period are acquired and converted into state diagnosis data, and the working state of a future expansion period is diagnosed through an intelligent diagnosis model. The working state of the main shaft system is self-adaptively adjusted; the working state of the spindle system at the future moment is diagnosed in advance in combination with the artificial intelligence technology, the problem of diagnosis result delay existing in real-time diagnosis operation is avoided, and the adjustment amplitude of abnormal characteristics can be effectively limited; adjustment errors caused by inconsistent adjustment amplitudes during synchronous adjustment of abnormal characteristics are avoided, and the synchronous working state and accurate optimization adjustment effect of the main shaft system are ensured.

Description

Technical Field

[0001] The present invention relates to the field of diagnosis and control technology, and more specifically, to an intelligent diagnosis system for abnormal working state of a spindle system of a five-axis machining center. Background Art

[0002] The five-axis machining center is the core equipment of high-end manufacturing equipment. The spindle system is the core functional module of the five-axis machining center, which is responsible for driving the tool to rotate at high speed and realize the cutting operation of the workpiece. Therefore, the health status of the spindle system directly affects the machining accuracy and equipment life of the five-axis machining center. In order to improve the machining efficiency of the five-axis machining center and avoid sudden failures of the spindle system, it is necessary to perform intelligent diagnosis of the working status of the spindle system.

[0003] The patent application with reference publication number CN115685878A discloses a fault diagnosis system based on the operation data of a five-axis CNC machine tool, including a data acquisition module, a data processing module, a data analysis module and a fault diagnosis module, wherein the data acquisition module is used to obtain the operation data of the five-axis CNC machine tool, the operation data including power data and posture data, the data processing module is used to process the power data and posture data of the five-axis CNC machine tool obtained by the data acquisition module, the data analysis module is used to analyze the power data and posture data of the five-axis CNC machine tool according to the processing results of the power data and posture data by the data processing module, and the fault diagnosis module is used to judge the cause of the abnormality; The existing intelligent diagnosis system for working status collects various parameters of the spindle system during operation in real time, combines the intelligent model to perform intelligent predictive diagnosis of the working status, and adjusts the abnormal parameters one by one. For example, in the above-mentioned patent application, it can judge and optimize the power operation status of the five-axis CNC machine tool according to the monitored power data in real time. However, when adjusting the abnormal parameters that affect the working status, the adjustment range of each abnormal parameter is usually controlled by the abnormal degree of the abnormal parameter, so that different abnormal parameters do not have a unified and reasonable adjustment standard when adjusting. When multiple abnormal parameters need to be adjusted synchronously, random and uncontrollable adjustment errors are prone to occur, thereby reducing the adjustment accuracy of the working status of the spindle system.

[0004] In view of this, the present invention proposes an intelligent diagnosis system for abnormal working status of a spindle system of a five-axis machining center to solve the above problem. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: an intelligent diagnosis system for abnormal working state of a spindle system of a five-axis machining center, comprising: The cycle calculation module is used to determine the cycle period of the spindle system in an effective state, collect the time limit change parameters of the spindle system in the cycle period, and plan the expansion cycle; The feature acquisition module is used to collect the comprehensive state characteristics of the spindle system during the expansion cycle, including temperature extremes, voltage values, current values, amplitude offset values, node jam values, and speed compliance rates, and convert the comprehensive state characteristics into state diagnosis data; The intelligent diagnosis module is used to perform intelligent diagnosis on the status diagnosis data through the trained intelligent diagnosis model, diagnose the working status of the future expansion cycle, and determine whether to execute the abnormal status adjustment mode; The state adjustment module is used to mark abnormal features in the state to be adjusted, and in combination with the adjustment control mechanism, adaptively adjust the working state of the spindle system. The adjustment control mechanism is: taking one adjustment unit as the adjustment standard, all abnormal features are adjusted synchronously.

[0006] Furthermore, the cycle period is: the moment when the spindle system generates the working parameters for the first time is taken as the starting moment, the current moment is taken as the ending moment, and the period between the starting moment and the ending moment is recorded as the cycle period.

[0007] Further, the term change parameter includes a unit temperature change value and a unit update value; The steps for collecting unit temperature change values ​​are as follows: At time T1 of the cycle period, the operating temperature of the spindle system is detected by a temperature sensor to obtain the initial temperature; The spindle system is started to run continuously. When the working temperature of the spindle system does not change within the preset temperature change time, the working temperature that has last changed is recorded as the termination temperature, and the time corresponding to the termination temperature is recorded as time T2; The absolute value of the difference between the initial temperature and the final temperature is taken, and compared with the time between time T1 and time T2 to obtain the unit temperature change value; The steps for collecting unit update values ​​are as follows: During the period, mark the database for data update Update events; Query by timestamp The update start time and update end time of each update event are recorded, and the duration between the update start time and the update end time is recorded as the update duration. After removing the maximum and minimum values ​​of the update duration, the remaining The update durations are accumulated and averaged to obtain the unit update value.

[0008] Furthermore, the planning steps for the expansion cycle are as follows: The unit temperature change value and the unit update value of the spindle system are added and averaged to obtain the expansion time. If the cycle period is an integer multiple of the expansion period, take one expansion period as the planning standard and plan the cycle period into B expansion periods; If the cycle period is not an integer multiple of the expansion time, take one tenth of the expansion time as the expansion value, and continuously increase the expansion value to the expansion time until the cycle period is an integer multiple of the increased expansion time, and take the increased expansion time as the planning standard to plan the cycle period into B expansion cycles.

[0009] Furthermore, the steps for collecting the amplitude offset value are as follows: In a stationary state, a stationary top view image of the spindle motor is captured by a camera, and a center point of the spindle motor in the stationary top view image is identified and recorded as a standard point; Mark all moments in B expansion cycles one by one to obtain C time points, use a camera to capture dynamic overhead images of the spindle motor at the C time points, and identify the center points of the spindle motor in the C dynamic overhead images one by one to obtain C dynamic points; Overlay the C dynamic overhead images one by one on the static overhead image, and measure the distances from the C dynamic points to the standard points one by one using a scale to obtain C sub-offset values; The sub-offset value greater than the calibrated offset value is recorded as the target offset value, and the target offset value, and The target offset values ​​are accumulated and averaged to obtain B amplitude offset values.

[0010] Furthermore, the steps for collecting node jam values ​​are as follows: The gyroscope installed on the spindle system monitors the direction of the tool in real time, and the position where the direction of movement changes is recorded as a node. All nodes in B expansion cycles are counted one by one, and the arrival time of all nodes is queried one by one through the timestamp; The control system of the five-axis machining center queries the three-dimensional coordinates of all nodes one by one, and records the time between the arrival times of two adjacent nodes as the total node time according to the order of arrival time, and obtains Total duration of each node; exist In the total duration of each node, the moment when the three-dimensional coordinates of the node change is recorded as the movement time, and after accumulating all the movement times, we get The movement duration of each node; Will The total duration of each node is compared with The movement time of each node is subtracted and the The differences are accumulated and averaged to obtain the jamming values ​​of B nodes.

[0011] Furthermore, the steps for collecting the speed compliance rate are as follows: The real-time speed of the transmission shaft at C time points is monitored in real time by a speed sensor, and the total duration of the time points when the real-time speed is greater than 0 is counted to obtain B total rotation durations; The time point when the real-time speed is greater than or equal to the target speed is recorded as the target time point, and the total duration of all target time points is calculated to obtain B target durations; After comparing the B target-reaching time durations with the B total rotation time durations one by one, the B rotation speed target-reaching rates are obtained.

[0012] Furthermore, the working state includes a normal state and an abnormal state; The training steps of the intelligent diagnosis model are as follows: Collect multiple groups of status diagnosis data of the spindle system in normal and abnormal states in advance; Convert multiple groups of status diagnosis data into multiple feature vectors using the sliding window method, convert the working status into labels corresponding to the status diagnosis data according to the sliding step, one feature vector corresponds to one label, and constitutes a group of training data. Multiple groups of training data constitute a training set, and preset the prediction time step, sliding step and sliding window length of the intelligent diagnosis model; The feature vector is used as the input of the intelligent diagnosis model, the working state of the future expansion cycle after the diagnosis time step is used as the output, the subsequent working state of each training set is used as the diagnosis target, and the sum of the minimized diagnosis errors is used as the training target to train the intelligent diagnosis model; The decision steps for executing the abnormal state adjustment mode are as follows: When the working state of the future expansion cycle is diagnosed as a normal state, it is determined that the abnormal state adjustment mode is not to be executed; When the diagnosed operating state of the future expansion cycle is an abnormal state, it is determined that the abnormal state adjustment mode is executed.

[0013] Furthermore, the method for identifying abnormal features is: Record the period value of the state diagnosis data whose working state is abnormal as the abnormal period value, and record the state diagnosis data corresponding to the period value located one digit above the abnormal period value as the adjustment data; Compare the temperature extreme value, voltage value, current value, amplitude offset value, node jam value and speed compliance rate in the adjustment data with the corresponding upper limit value or lower limit value one by one; When the temperature extreme value is greater than the upper temperature limit, the temperature extreme value is recorded as an abnormal feature; When the voltage value is greater than the voltage upper limit or less than the voltage lower limit, the voltage value is recorded as an abnormal feature; When the current value is greater than the upper current limit or less than the lower current limit, the current value is recorded as an abnormal feature; When the amplitude offset value is greater than the offset upper limit value, the amplitude offset value is recorded as an abnormal feature; When the node jam value is greater than the jam upper limit, the node jam value is recorded as an abnormal feature; When the speed compliance rate is less than the compliance upper limit, the speed compliance rate is recorded as an abnormal feature.

[0014] Furthermore, the steps of adaptively adjusting the working state are as follows: A01: Subtract all abnormal features from the corresponding upper or lower limit values ​​one by one and take the absolute value to obtain the abnormal difference, count the number of abnormal features, and compare the minimum value of the abnormal difference with the number of abnormal features to obtain the adjustment unit; A02: Taking one adjustment unit as the adjustment standard, all abnormal features in the adjustment data are increased or decreased simultaneously, and the adjusted status diagnosis data is input into the intelligent diagnosis model; A03: When the working state of the future expansion cycle is diagnosed as abnormal, the above A02 is repeated cyclically until the working state of the future expansion cycle is diagnosed as normal, and the adjustment is stopped to complete the adaptive adjustment operation of the working state of the spindle system.

[0015] The technical effects and advantages of the intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center of the present invention are as follows: (1) By determining the cycle period of the spindle system and planning the expansion period based on the unit temperature change value and the unit update value, the relevant parameters of the basic operating status of the spindle system within the cycle period can be accurately collected to ensure that the planning of subsequent expansion cycles can be closely related to the real-time operation of the spindle system, so that the expansion cycle can evenly divide the cycle period, providing a precise, reasonable and accurate time standard for the subsequent intelligent diagnosis of the working status of the spindle system.

[0016] (2) By collecting the comprehensive state characteristics of the spindle system during the expansion cycle and converting the comprehensive state characteristics into state diagnosis data, the comprehensive state characteristics can accurately represent the characteristic parameters that affect whether the working state of the spindle system is abnormal or not, avoiding the limitations caused by single-dimensional or one-sided feature collection, and thus can have a comprehensive and diversified collection effect on the working state of the spindle system. The working state of the future expansion cycle can be diagnosed through the intelligent diagnosis model, and it can be determined whether to execute the abnormal state adjustment mode. It can be combined with artificial intelligence technology to make an early diagnosis of whether the working state of the spindle system at a future moment is abnormal, thereby realizing the intelligent diagnosis effect of the working state of the spindle system beyond the timeline, avoiding the problem of delayed diagnosis results in real-time detection and diagnosis operations.

[0017] (3) By marking abnormal features and combining the adjustment control mechanism, the working state of the spindle system can be adaptively adjusted. The single adjustment range of the abnormal features can be effectively limited in combination with the limiting effect of the adjustment control mechanism, thereby avoiding the adjustment error caused by inconsistent adjustment ranges when multiple abnormal features are adjusted synchronously, thereby ensuring that the working state of the spindle system can be optimized and adjusted synchronously and accurately, and achieving the adaptive adjustment effect of the working state of the spindle system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of a module of an intelligent diagnosis system for abnormal working state of a spindle system of a five-axis machining center provided in the first embodiment of the present invention; Figure 2 A flowchart of an intelligent diagnosis method for abnormal working status of a spindle system of a five-axis machining center provided in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Example 1: Please refer to Figure 1 As shown, the intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center described in this embodiment includes: The cycle calculation module, in an effective state, determines the cycle period of the spindle system, collects the time limit change parameters of the spindle system within the cycle period, and plans the expansion cycle; The spindle system is the core functional module of the five-axis machining center, which is responsible for driving the tool to rotate at high speed and realize the cutting operation of the workpiece. The spindle system usually includes a spindle power supply, a spindle motor, a spindle transmission structure and a spindle tool assembly.

[0021] The effective state refers to the state in which the working environment of the spindle system of the five-axis machining center is in safe operation. By judging whether the spindle system is in an effective state, the current working environment of the spindle system can be preliminarily identified. Therefore, the quality of the working environment will directly affect whether the spindle system is in an effective state. In order to determine whether the spindle system is in a valid state, it is necessary to identify the actual working environment of the five-axis machining center. Therefore, the state in which the working environment temperature of the five-axis machining center is within the safe ambient temperature range and the working environment humidity is within the safe ambient humidity range is recorded as a valid state.

[0022] The above-mentioned safe ambient temperature range and safe ambient humidity range refer to the external ambient temperature range and external ambient humidity range in which the five-axis machining center can operate normally and effectively. The safe ambient temperature range and the safe ambient humidity range are obtained by querying the technical safety manual of the five-axis machining center.

[0023] When the five-axis machining center is in an effective state, it is necessary to determine the cycle period of the spindle system, so that the cycle period can be used as the overall time span for the intelligent diagnosis of whether the working state of the spindle system is abnormal, and all working parameters of the spindle system within the cycle period can be used as the basis for affecting the working state, and the working parameters of the spindle system include but are not limited to temperature parameters, current parameters, voltage parameters, speed parameters, displacement parameters, etc.; Therefore, the time span corresponding to the cycle period is large, and the number of parameters included is also large. The cycle period is: the moment when the spindle system first generates working parameters is the starting moment, the current moment is the ending moment, and the period between the starting moment and the ending moment is recorded as the cycle period.

[0024] After the cycle period is obtained, the time limit variation parameters of the spindle system within the cycle period can be collected, so that the time limit variation parameters can affect the duration of the spindle system working parameters changing once within the cycle period and the change amplitude is sufficient to be collected and identified; The term change parameters include unit temperature change value and unit update value; the unit temperature change value refers to the increase or decrease amplitude of the working temperature of the spindle system in unit time. The larger the unit temperature change value, the larger the increase or decrease amplitude of the working temperature of the spindle system in unit time. The steps for collecting unit temperature change values ​​are as follows: At time T1 of the cycle period, the operating temperature of the spindle system is detected by a temperature sensor to obtain an initial temperature; Start the spindle system to run continuously. When the working temperature of the spindle system does not change within the preset temperature change time, the last changed working temperature is recorded as the termination temperature, and the time corresponding to the termination temperature is recorded as the T2 time. The preset temperature change time refers to the maximum time corresponding to the temperature change of a magnitude sufficient to be collected, which can ensure that the working temperature of the spindle system can change within the preset temperature change time. If the working temperature does not change within the preset temperature change time, it means that the spindle system is in a shutdown state. The absolute value of the difference between the initial temperature and the final temperature is taken, and compared with the time between time T1 and time T2 to obtain the unit temperature change value; The calculation formula for unit temperature change is: ; In the formula, is the unit temperature change value, is the initial temperature, is the termination temperature, It is the duration between time T1 and time T2.

[0025] The unit update value refers to the time it takes for the spindle system's working parameters to update data once. The larger the unit update value, the longer it takes for the spindle system's working parameters to update data once. The steps for collecting unit update values ​​are as follows: During the cycle period, mark the events of data update in the database of the five-axis machining center to obtain Update events; Query by timestamp The update start time and update end time of each update event are recorded, and the duration between the update start time and the update end time is recorded as the update duration. After removing the maximum and minimum values ​​of the update duration, the remaining The update durations are accumulated and averaged to obtain the unit update value; The calculation formula for the unit update value is; ; In the formula, Update the value for the unit, For the Update duration.

[0026] The expansion cycle refers to the time segment after the period of the spindle system is finely divided, so that the expansion cycle can divide the larger span duration corresponding to the period into smaller span durations, and serve as the time basis for subsequent intelligent diagnosis of whether the working status of the spindle system is abnormal; The steps for planning a bulking cycle are as follows: The unit temperature change value and the unit update value of the spindle system are added and averaged to obtain the expansion time. The expansion time is calculated as: ; In the formula, is the expansion duration; If the cycle time period is an integer multiple of the expansion time length, then one expansion time length is used as the planning standard, and the cycle time period is planned into B expansion cycles; if it is an integer multiple, it means that the expansion time length can evenly divide the cycle time period, otherwise it is the opposite; If the cycle period is not an integer multiple of the expansion time, then one tenth of the expansion time is used as the expansion value, and an expansion value is continuously added to the expansion time until the cycle period is an integer multiple of the increased expansion time, and the increased expansion time is used as the planning standard to plan the cycle period into B expansion cycles.

[0027] It should be noted that when the expansion time is increased, the final increase result is less than or equal to the maximum value of the unit temperature change value and the unit update value, which can ensure that the increased expansion time will not be greater than any of the unit temperature change value and the unit update value, thereby achieving the effect of equal time division of the cycle time periods.

[0028] The feature acquisition module collects the comprehensive state features of the spindle system in the expansion cycle in turn and converts the comprehensive state features into state diagnosis data. The comprehensive state features include temperature extreme value, voltage value, current value, amplitude offset value, node jam value and speed compliance rate; After planning the expansion cycle of the spindle system, it is necessary to collect the comprehensive state characteristics of the spindle system in the expansion cycle, so that the comprehensive state characteristics can be used as specific data of the working state of the spindle system in each expansion cycle, and as data support for judging whether the specific working state of the spindle system in each expansion cycle is abnormal or not; The comprehensive status characteristics include temperature extreme value, voltage value, current value, amplitude offset value, node jam value and speed compliance rate; The temperature extreme value refers to the maximum value of the tool surface temperature of the spindle system during the expansion cycle, which can be used to indicate the high or low temperature of the tool surface of the spindle system. The larger the temperature extreme value, the worse the heat dissipation and cooling performance of the tool of the spindle system during cutting work, and the greater the probability of abnormal working state of the spindle system. The temperature extreme value is obtained by measuring the tool surface temperature in B expansion cycles in real time through an infrared temperature measuring device installed on the five-axis machining center, and taking the maximum value.

[0029] The voltage value and current value refer to the average voltage and average current of the power supply of the spindle system during the expansion cycle, respectively, which can represent the stability of the power supply voltage and power supply current of the spindle system; the voltage value and current value are obtained by real-time detecting the minimum and maximum values ​​of the voltage of the power supply of the spindle system during B expansion cycles and the minimum and maximum values ​​of the real-time current through a voltage sensor and a current sensor, and then adding the maximum and minimum values ​​of the real-time voltage and real-time current and averaging them.

[0030] The amplitude offset value refers to the amplitude of the mechanical vibration offset of the spindle motor of the spindle system during the expansion period, which can be used to indicate the stability of the spindle motor of the spindle system. The larger the amplitude offset value, the larger the amplitude of the mechanical vibration offset of the spindle motor of the spindle system during the expansion period, and the greater the probability of abnormal working state of the spindle system. The steps for collecting the amplitude offset value are as follows: In a stationary state, a camera is used to capture a stationary top-view image of the spindle motor, and the center point of the spindle motor in the stationary top-view image is identified and recorded as a standard point; the center point refers to the middle point of the spindle motor itself, which is located and identified by computer vision technology; Mark all moments in B expansion cycles one by one to obtain C time points, use a camera to capture dynamic overhead images of the spindle motor at the C time points, and identify the center points of the spindle motor in the C dynamic overhead images one by one to obtain C dynamic points; Overlay the C dynamic overhead images one by one on the static overhead image, and measure the distances from the C dynamic points to the standard points one by one using a scale to obtain C sub-offset values; The sub-offset value greater than the calibrated offset value is recorded as the target offset value, and the target offset value, and The target offset values ​​are accumulated and averaged to obtain B amplitude offset values; the calibration offset value is used to numerically represent the vibration effect on the spindle motor during the operation of the five-axis machining center itself, so that the sub-offset value exceeding the calibration offset value will be recorded as the abnormal vibration offset of the spindle motor itself; The amplitude offset value is calculated as:

[0032] In the formula, For the The amplitude deviation value of the expansion cycle is For the The expansion cycle target offset value.

[0033] The node jam value refers to the pause time of the tool of the spindle system from moving to a node in one direction to starting to move to a node in another direction. It can be used to indicate the reaction performance of the tool of the spindle system. The larger the node jam value, the worse the reaction performance of the tool of the spindle system, and the greater the probability of abnormal phenomenon in the working state of the spindle system. The steps for collecting node jam values ​​are as follows: The gyroscope installed on the spindle system monitors the direction of the tool in real time, and the position where the direction of movement changes is recorded as a node. All nodes in B expansion cycles are counted one by one, and the arrival time of all nodes is queried one by one through the timestamp; The control system of the five-axis machining center queries the three-dimensional coordinates of all nodes one by one, and records the time between the arrival times of two adjacent nodes as the total node time according to the order of arrival time, and obtains Total duration of each node; exist In the total duration of each node, the moment when the three-dimensional coordinates of the node change is recorded as the movement time, and after accumulating all the movement times, we get The movement duration of each node; Will The total duration of each node is compared with The movement time of each node is subtracted and the The differences are accumulated and averaged to obtain the jamming values ​​of B nodes; The calculation formula for the node jam value is: ; In the formula, For the The node lag value of expansion cycles, For the The expansion cycle The total duration of each node, For the The expansion cycle The movement duration of each node.

[0034] The speed compliance rate refers to the proportion of time that the speed of the transmission shaft of the spindle system is in the compliance state during the expansion cycle, which can be used to indicate the speed compliance performance of the transmission shaft of the spindle system. The greater the speed compliance rate, the greater the proportion of time that the speed of the transmission shaft of the spindle system is in the compliance state during the expansion cycle, and the smaller the probability of abnormal phenomena in the working state of the spindle system. The steps for collecting the speed compliance rate are as follows: The real-time speed of the transmission shaft at C time points is monitored in real time by a speed sensor, and the total duration of the time points when the real-time speed is greater than 0 is counted to obtain B total rotation durations; The time point when the real-time speed is greater than or equal to the target speed is recorded as the target time point, and the total duration of all target time points is calculated to obtain B target durations; the target speed refers to the lowest speed of the transmission shaft during normal processing operation, thereby avoiding the low speed phenomenon of the transmission shaft in the early start time period and the late stop time period; After comparing the B target-reaching durations one by one with the B total rotation durations, the target-reaching rates of the B rotation speeds are obtained; The calculation formula of the speed compliance rate is: ; In the formula, For the The speed reaching rate of each expansion cycle, For the The duration of the expansion cycle is No. The total rotation time of an expansion cycle.

[0035] After obtaining the comprehensive state characteristics of all expansion cycles, it is now necessary to transform all the comprehensive state characteristics, so as to effectively arrange and transform the comprehensive state characteristics in a discrete and irregular state, so that the comprehensive state characteristics can be transformed into state diagnosis data; When converting the status diagnostic data, it is necessary to use the time sequence corresponding to the B expansion cycles in the cycle time period as the basis, assign cycle values ​​to the B expansion cycles in turn, and summarize and combine the comprehensive status characteristics of the B expansion cycles to obtain B data combinations, and then annotate the cycle values ​​of the B expansion cycles on the B data combinations to obtain B status diagnostic data.

[0036] It should be noted that the purpose of assigning cycle values ​​to B expansion cycles is to accurately distinguish the B expansion cycles and to serve as a basis for determining the order of the acquisition time of the B expansion cycles. For example, the cycle values ​​assigned to the B expansion cycles are recorded as 0B1, 0B2, 0B3, ..., 0BB, respectively. After the temperature extremes, voltage values, current values, amplitude offset values, node jam values ​​and speed compliance rates within the B expansion cycles are summarized and combined to obtain B data combinations, 0B1, 0B2, 0B3, ..., 0BB are respectively noted on the B data combinations to obtain B status diagnostic data.

[0037] The intelligent diagnosis module uses the trained intelligent diagnosis model to perform intelligent diagnosis on the status diagnosis data, diagnose the working status of the future expansion cycle, and determine whether to execute the abnormal status adjustment mode; The intelligent diagnosis model is a machine learning model that uses the collected status diagnosis data as a basis and combines artificial intelligence technology to perform intelligent diagnosis and analysis on the working status of the spindle system in different expansion cycles. This allows for advanced diagnosis and processing of the working status of the spindle system on the timeline, and the determination of the working status of the spindle system in future expansion cycles.

[0038] The working status is used to indicate whether the specific cutting processing operation status corresponding to the spindle system during the expansion cycle is abnormal or not. The working status includes normal status and abnormal status. The normal status means that the specific cutting processing operation process status corresponding to the spindle system during the expansion cycle is normal and fault-free. The abnormal status means that the specific cutting processing operation process status corresponding to the spindle system during the expansion cycle is abnormal and faulty. The normal status and abnormal status are obtained by querying the actual status corresponding to the status diagnosis data in different expansion cycles through the control terminal of the five-axis machining center.

[0039] The training steps of the intelligent diagnosis model are as follows: Collect multiple groups of status diagnosis data of the spindle system in normal and abnormal states in advance; Convert multiple groups of status diagnosis data into multiple feature vectors using a sliding window method, convert the working state into a label corresponding to the status diagnosis data according to the sliding step, convert the normal state and the abnormal state into digital labels respectively, exemplarily, convert the normal state into 0, and convert the abnormal state into 1, one feature vector corresponds to one label, and constitutes a group of training data, and multiple groups of training data constitute a training set, arrange the status diagnosis data in order from small to large according to the period value, and preset the prediction time step, sliding step and sliding window length of the intelligent diagnosis model; The feature vector is used as the input of the intelligent diagnosis model, the working state of the future expansion cycle after the diagnosis time step is used as the output, the subsequent working state of each training set is used as the diagnosis target, and the sum of the minimized diagnosis errors is used as the training target. The intelligent diagnosis model is trained to generate an intelligent diagnosis model that can diagnose the working state of the future expansion cycle based on the state diagnosis data.

[0040] After the intelligent diagnosis model is trained, the status diagnosis data can be input into the intelligent diagnosis model, so that the intelligent diagnosis model can predict the working status of the future expansion cycle according to the input intelligent diagnosis model; Specifically, when the output of the intelligent diagnosis model is 0, it means that the working state of the diagnosed spindle system in the future expansion cycle will not have abnormal faults, and the working state of the future expansion cycle is normal; When the output of the intelligent diagnosis model is 1, it indicates that the working state of the diagnosed spindle system in the future expansion cycle will have an abnormal fault phenomenon, and the working state of the future expansion cycle is an abnormal state.

[0041] After the intelligent diagnosis model diagnoses the working state of the spindle system in the future expansion cycle, it can be determined whether it is necessary to execute the abnormal state adjustment mode before the spindle system has an abnormal working state in the future expansion cycle according to the actual diagnosis result. The abnormal state adjustment mode is a working mode for making targeted matching adjustments to whether the working state of the spindle system is abnormal or not. This mode will only be executed when the working state of the spindle system in the future expansion cycle is abnormal. The decision steps for executing the abnormal state adjustment mode are as follows: When the working state of the future expansion cycle is diagnosed as a normal state, it is not necessary to optimize and adjust the working state of the spindle system, and it is determined that the abnormal state adjustment mode is not executed; When the working state of the future expansion cycle is diagnosed as an abnormal state, it is necessary to optimize and adjust the working state of the spindle system, and it is determined to execute the abnormal state adjustment mode.

[0042] The state adjustment module marks the abnormal features in the state to be adjusted, and combines the adjustment control mechanism to adaptively adjust the working state of the spindle system; When the abnormal state adjustment mode is executed, it indicates that the working state of the spindle system in the future expansion cycle is prone to abnormal failure. At this time, it is necessary to further identify the state diagnosis data of the spindle system in the expansion cycle before the abnormal failure occurs in the working state of the spindle system, so as to record the specific comprehensive state characteristics of the abnormal phenomenon after identification as abnormal characteristics, and the abnormal characteristics at this time will be in a state to be adjusted and serve as the direct object for subsequent adjustment of the working state of the spindle system; The method for identifying abnormal features is: Record the period value of the state diagnosis data whose working state is abnormal as the abnormal period value, and record the state diagnosis data corresponding to the period value located one digit above the abnormal period value as the adjustment data; Compare the temperature extreme value, voltage value, current value, amplitude offset value, node jam value and speed compliance rate in the adjustment data with the corresponding upper limit value or lower limit value one by one; When the temperature extreme value is greater than the temperature upper limit value, it means that the maximum temperature of the tool surface of the spindle system exceeds the upper limit of the safe temperature, and the temperature extreme value is recorded as an abnormal feature; the temperature upper limit value refers to the maximum value when the temperature extreme value is not recorded as an abnormal feature; When the voltage value is greater than the voltage upper limit value or less than the voltage lower limit value, it means that the voltage of the power supply of the spindle system is not within the safe range, and the voltage value is recorded as an abnormal feature; the voltage upper limit value and the voltage lower limit value refer to the maximum value and the minimum value when the voltage value is not recorded as an abnormal feature; When the current value is greater than the current upper limit value or less than the current lower limit value, it means that the current of the power supply of the spindle system is not within the safe range, and the current value is recorded as an abnormal feature; the current upper limit value and the current lower limit value refer to the maximum value and the minimum value when the current value is not recorded as an abnormal feature; When the amplitude offset value is greater than the offset upper limit value, it means that the vibration offset of the spindle motor of the spindle system exceeds the upper limit of the safety offset, and the amplitude offset value is recorded as an abnormal feature; the offset upper limit value refers to the maximum value when the amplitude offset value is not recorded as an abnormal feature; When the node jam value is greater than the jam upper limit, it means that the jam duration of the spindle system exceeds the upper limit of the safe jam, and the node jam value is recorded as an abnormal feature; the jam upper limit refers to the maximum value when the node jam value is not recorded as an abnormal feature; When the speed compliance rate is less than the compliance upper limit, it means that the speed compliance time of the spindle system is lower than the lower limit of the safety compliance time, and the speed compliance rate is recorded as an abnormal feature. The compliance upper limit refers to the minimum value when the speed compliance rate is not recorded as an abnormal feature.

[0043] After identifying abnormal features from the comprehensive state features, these abnormal features can be used as the object of subsequent optimization and adjustment of the working state of the spindle system, thereby preventing the spindle system of the five-axis machining center from working abnormally in the future expansion cycle and avoiding the spindle system from operating failures at any time in the future; When optimizing and adjusting the working state of the spindle system, it is necessary to optimize and adjust according to the adjustment control mechanism to ensure that the spindle system can reduce the number and steps of optimization adjustments of abnormal characteristics of the spindle system as much as possible without abnormal working state in the future expansion cycle, so as to achieve the adaptive dynamic adjustment effect of the spindle system; The adjustment control mechanism is: take one adjustment unit as the adjustment standard to adjust all abnormal features synchronously; the adjustment unit refers to the adjustment amplitude of increasing or decreasing the value of the abnormal feature, so as to ensure that each abnormal feature can be adjusted according to the fixed adjustment amplitude; The steps for adaptive adjustment of working status are as follows: A01: Subtract all abnormal features from the corresponding upper or lower limit values ​​one by one and take the absolute value to obtain the abnormal difference, count the number of abnormal features, and compare the minimum value of the abnormal difference with the number of abnormal features to obtain the adjustment unit; The calculation formula for the adjustment unit is: ; In the formula, To adjust the unit, is the minimum value of the abnormal difference, is the number of abnormal features; A02: Taking one adjustment unit as the adjustment standard, all abnormal features in the adjustment data are increased or decreased simultaneously, and the adjusted status diagnosis data is input into the intelligent diagnosis model; A03: When the working state of the future expansion cycle is diagnosed as abnormal, the above A02 is repeated cyclically until the working state of the future expansion cycle is diagnosed as normal, and the adjustment is stopped to complete the adaptive adjustment operation of the working state of the spindle system.

[0044] Specifically, if the abnormal characteristics are temperature extremes, voltage values, current values, amplitude offset values, node jamming values, and speed compliance rates, respectively, when adjusting the working state of the spindle system, it is necessary to first subtract the temperature extremes, voltage values, current values, amplitude offset values, node jamming values, and speed compliance rates from the corresponding upper limits, and compare the minimum value of the abnormal difference with the number of abnormal characteristics to obtain an adjustment unit, and then use one adjustment unit as the amplitude of one adjustment to adjust the size of all abnormal characteristics. For example, the abnormal characteristic corresponding to the increase adjustment operation is the speed compliance rate, and the abnormal characteristic corresponding to the decrease adjustment operation is the temperature extremes, amplitude offset values, and node jamming values. When the voltage value and the current value are greater than the corresponding upper limits, it is necessary to reduce the voltage value and the current value, and vice versa. This can achieve the adaptive dynamic adjustment effect of the abnormal characteristics that cause abnormal phenomena in the working state of the spindle system, ensuring that the spindle system of the five-axis machining center can always maintain a normal working state.

[0045] In this embodiment, by determining the cycle time period of the spindle system and planning the expansion cycle based on the unit temperature change value and the unit update value, the relevant parameters of the basic operating state of the spindle system within the cycle time period can be accurately collected to ensure that the planning of subsequent expansion cycles can be closely related to the real-time operation of the spindle system, so that the expansion cycle can evenly divide the cycle time period, providing a precise, reasonable and accurate duration standard for the subsequent intelligent diagnosis of the working state of the spindle system.

[0046] By collecting the comprehensive state characteristics of the spindle system during the expansion cycle and converting the comprehensive state characteristics into state diagnosis data, the comprehensive state characteristics can accurately represent the characteristic parameters that affect whether the working state of the spindle system is abnormal or not, avoiding the limitations caused by single-dimensional or one-sided feature collection, and thus can have a comprehensive and diversified collection effect on the working state of the spindle system. The working state of the future expansion cycle can be diagnosed through an intelligent diagnosis model, and it can be determined whether to execute the abnormal state adjustment mode. Artificial intelligence technology can be combined to diagnose in advance whether the working state of the spindle system at a future moment is abnormal or not, thereby realizing the intelligent diagnosis effect of the working state of the spindle system beyond the timeline, and avoiding the problem of delayed diagnosis results in real-time detection and diagnosis operations.

[0047] By marking abnormal features and combining them with the adjustment control mechanism, the working state of the spindle system can be adaptively adjusted. The single adjustment range of the abnormal features can be effectively limited in combination with the limiting effect of the adjustment control mechanism, thereby avoiding adjustment errors caused by inconsistent adjustment ranges when multiple abnormal features are adjusted synchronously. This ensures that the working state of the spindle system can be optimized and adjusted synchronously and accurately, thereby achieving an adaptive adjustment effect of the working state of the spindle system.

[0048] Specifically, the time when the spindle system first generates working parameters is 09:08:22 on February 3, 2025, and the current time is 09:13:52 on February 3, 2025, then the cycle period is 330 seconds, the database automatically collects the unit temperature change value and unit update value of the spindle system, and calculates that the expansion time is 5 seconds, at this time, the cycle period can be divided into 66 expansion cycles; The 66 expansion cycles are assigned cycle values, which are 0B1, 0B2, ..., 0B66 respectively. In the first expansion cycle, the temperature extreme value is 60°C, the voltage value is 220V, the current value is 10A, the amplitude offset value is 0.05mm, the node jam value is 0.20s, and the speed compliance rate is 96.5%. Then the first state diagnosis data is 0B1, 60°C, 220V, 10A, 0.05mm, 0.20s and 96.5%; and so on and so forth until the state diagnosis data corresponding to all 66 expansion cycles are obtained, and the working states of the spindle system corresponding to the 66 state diagnosis data are collected one by one and matched with them, so as to achieve the association matching effect between the 66 state diagnosis data and the working state, and at the same time, the 66 state diagnosis data and the working state are summarized and sorted to form a cycle state table, which is specifically shown as follows:

[0049] It can be seen from the above cycle state table that within the planned 66 expansion cycles, each expansion cycle corresponds to a unique temperature extreme value, voltage value, current value, amplitude offset value, node jam value, speed compliance rate and working status. Therefore, all the data in each column of the cycle state table can be summarized to generate status diagnostic data, which can be used to train the intelligent diagnosis model, thereby realizing the intelligent diagnosis effect of the working status of the spindle system of the five-axis machining center in the future expansion cycle.

[0050] Example 2: Please refer to Figure 2 As shown, the part not described in detail in this embodiment is described in the first embodiment, and a smart diagnosis method for abnormal working state of a spindle system of a five-axis machining center is provided, which is implemented based on a smart diagnosis system for abnormal working state of a spindle system of a five-axis machining center, and includes: S1: In the effective state, determine the period of the spindle system, collect the period change parameters of the spindle system in the period, and plan the expansion period; The cycle period is: the time when the spindle system first generates working parameters is the starting time, the current time is the ending time, and the period from the starting time to the ending time is recorded as the cycle period. The period change parameters include the unit temperature change value and the unit update value; S2: Collecting comprehensive state characteristics of the spindle system during the expansion cycle, and converting the comprehensive state characteristics into state diagnosis data; The comprehensive status characteristics include temperature extreme value, voltage value, current value, amplitude offset value, node jam value and speed compliance rate; When converting the state diagnosis data, the time sequence of the B expansion cycles in the cycle period is taken as the basis, the B expansion cycles are assigned cycle values ​​in turn, and the comprehensive state characteristics of the B expansion cycles are summarized and combined to obtain B data combinations, and then the cycle values ​​of the B expansion cycles are correspondingly annotated on the B data combinations to obtain B state diagnosis data; S3: Perform intelligent diagnosis on the status diagnosis data through the trained intelligent diagnosis model, diagnose the working status of the future expansion cycle, and determine whether to execute the abnormal status adjustment mode; The working state includes normal state and abnormal state; S4: if the abnormal state adjustment mode is executed, the abnormal features in the state to be adjusted are marked, and the working state of the spindle system is adaptively adjusted in combination with the adjustment control mechanism; The adjustment control mechanism is: taking one adjustment unit as the adjustment standard, all abnormal features are adjusted synchronously.

[0051] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An intelligent diagnosis system for abnormal working status of the spindle system of a five-axis machining center, characterized in that: include: The cycle calculation module is used to determine the cycle period of the spindle system in an effective state, collect the time limit change parameters of the spindle system in the cycle period, and plan the expansion cycle; The feature acquisition module is used to collect the comprehensive state characteristics of the spindle system during the expansion cycle, including temperature extremes, voltage values, current values, amplitude offset values, node jam values, and speed compliance rates, and convert the comprehensive state characteristics into state diagnosis data; The intelligent diagnosis module is used to perform intelligent diagnosis on the status diagnosis data through the trained intelligent diagnosis model, diagnose the working status of the future expansion cycle, and determine whether to execute the abnormal status adjustment mode; The state adjustment module is used to mark abnormal features in the state to be adjusted, and in combination with the adjustment control mechanism, adaptively adjust the working state of the spindle system. The adjustment control mechanism is: taking one adjustment unit as the adjustment standard, all abnormal features are adjusted synchronously.

2. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 1 is characterized in that: The cycle period is: the time when the spindle system generates working parameters for the first time is the starting time, the current time is the ending time, and the period from the starting time to the ending time is recorded as the cycle period.

3. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 2 is characterized in that: The term change parameters include unit temperature change value and unit update value; The steps for collecting unit temperature change values ​​are as follows: At time T1 of the cycle period, the operating temperature of the spindle system is detected by a temperature sensor to obtain an initial temperature; The spindle system is started to run continuously. When the working temperature of the spindle system does not change within the preset temperature change time, the working temperature that has last changed is recorded as the termination temperature, and the time corresponding to the termination temperature is recorded as time T2; The absolute value of the difference between the initial temperature and the final temperature is taken, and compared with the time between time T1 and time T2 to obtain the unit temperature change value; The steps for collecting unit update values ​​are as follows: During the period, mark the database for data update Update events; Query by timestamp The update start time and update end time of each update event are recorded, and the duration between the update start time and the update end time is recorded as the update duration. After removing the maximum and minimum values ​​of the update duration, the remaining The update durations are accumulated and averaged to obtain the unit update value.

4. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 3 is characterized in that: The steps for planning a bulking cycle are as follows: The unit temperature change value and the unit update value of the spindle system are added and averaged to obtain the expansion time. If the cycle period is an integer multiple of the expansion period, take one expansion period as the planning standard and plan the cycle period into B expansion periods; If the cycle period is not an integer multiple of the expansion time, take one tenth of the expansion time as the expansion value, and continuously increase the expansion value to the expansion time until the cycle period is an integer multiple of the increased expansion time, and take the increased expansion time as the planning standard to plan the cycle period into B expansion cycles.

5. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 4 is characterized in that: The steps for collecting the amplitude offset value are as follows: In a stationary state, a stationary top view image of the spindle motor is captured by a camera, and a center point of the spindle motor in the stationary top view image is identified and recorded as a standard point; Mark all moments in B expansion cycles one by one to obtain C time points, use a camera to capture dynamic overhead images of the spindle motor at the C time points, and identify the center points of the spindle motor in the C dynamic overhead images one by one to obtain C dynamic points; Overlay the C dynamic overhead images one by one on the static overhead image, and measure the distances from the C dynamic points to the standard points one by one using a scale to obtain C sub-offset values; The sub-offset value greater than the calibrated offset value is recorded as the target offset value, and the target offset value, and The target offset values ​​are accumulated and averaged to obtain B amplitude offset values.

6. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 5 is characterized in that: The steps for collecting node jam values ​​are as follows: The gyroscope installed on the spindle system monitors the direction of the tool in real time, and the position where the direction of movement changes is recorded as a node. All nodes in B expansion cycles are counted one by one, and the arrival time of all nodes is queried one by one through the timestamp; The control system of the five-axis machining center queries the three-dimensional coordinates of all nodes one by one, and records the time between the arrival times of two adjacent nodes as the total node time according to the order of arrival time, and obtains Total duration of each node; exist In the total duration of each node, the moment when the three-dimensional coordinates of the node change is recorded as the movement time, and after accumulating all the movement times, we get The movement duration of each node; Will The total duration of each node is compared with The movement time of each node is subtracted and the The differences are accumulated and averaged to obtain the jamming values ​​of B nodes.

7. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 6 is characterized in that: The steps for collecting the speed compliance rate are as follows: The real-time speed of the transmission shaft at C time points is monitored in real time by a speed sensor, and the total duration of the time points when the real-time speed is greater than 0 is counted to obtain B total rotation durations; The time point when the real-time speed is greater than or equal to the target speed is recorded as the target time point, and the total duration of all target time points is calculated to obtain B target durations; After comparing the B target-reaching time durations with the B total rotation time durations one by one, the B rotation speed target-reaching rates are obtained.

8. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 7 is characterized in that: The working state includes normal state and abnormal state; The training steps of the intelligent diagnosis model are as follows: Collect multiple groups of status diagnosis data of the spindle system in normal and abnormal states in advance; Convert multiple groups of status diagnosis data into multiple feature vectors using the sliding window method, convert the working status into labels corresponding to the status diagnosis data according to the sliding step, one feature vector corresponds to one label, and constitutes a group of training data. Multiple groups of training data constitute a training set, and preset the prediction time step, sliding step and sliding window length of the intelligent diagnosis model; The feature vector is used as the input of the intelligent diagnosis model, the working state of the future expansion cycle after the diagnosis time step is used as the output, the subsequent working state of each training set is used as the diagnosis target, and the sum of the minimized diagnosis errors is used as the training target to train the intelligent diagnosis model; The decision steps for executing the abnormal state adjustment mode are as follows: When the working state of the future expansion cycle is diagnosed as a normal state, it is determined that the abnormal state adjustment mode is not to be executed; When the diagnosed operating state of the future expansion cycle is an abnormal state, it is determined that the abnormal state adjustment mode is executed.

9. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 8, characterized in that: The method for identifying abnormal features is: Record the period value of the state diagnosis data whose working state is abnormal as the abnormal period value, and record the state diagnosis data corresponding to the period value located one digit above the abnormal period value as the adjustment data; Compare the temperature extreme value, voltage value, current value, amplitude offset value, node jam value and speed compliance rate in the adjustment data with the corresponding upper limit value or lower limit value one by one; When the temperature extreme value is greater than the upper temperature limit, the temperature extreme value is recorded as an abnormal feature; When the voltage value is greater than the voltage upper limit or less than the voltage lower limit, the voltage value is recorded as an abnormal feature; When the current value is greater than the upper current limit or less than the lower current limit, the current value is recorded as an abnormal feature; When the amplitude offset value is greater than the offset upper limit value, the amplitude offset value is recorded as an abnormal feature; When the node jam value is greater than the jam upper limit, the node jam value is recorded as an abnormal feature; When the speed compliance rate is less than the compliance upper limit, the speed compliance rate is recorded as an abnormal feature.

10. The intelligent diagnosis system for abnormal working state of the spindle system of a five-axis machining center according to claim 9, characterized in that: The steps for adaptive adjustment of working status are as follows: A01: Subtract all abnormal features from the corresponding upper or lower limit values ​​one by one and take the absolute value to obtain the abnormal difference, count the number of abnormal features, and compare the minimum value of the abnormal difference with the number of abnormal features to obtain the adjustment unit; A02: Taking one adjustment unit as the adjustment standard, all abnormal features in the adjustment data are increased or decreased simultaneously, and the adjusted status diagnosis data is input into the intelligent diagnosis model; A03: When the working state of the future expansion cycle is diagnosed as abnormal, the above A02 is repeated cyclically until the working state of the future expansion cycle is diagnosed as normal, and the adjustment is stopped to complete the adaptive adjustment operation of the working state of the spindle system.

Images