Intelligent gantry machining center fault monitoring method and system

Abstract

The invention relates to the technical field of machine tool operation and maintenance, in particular to an intelligent gantry machining center fault monitoring method and system, and the method comprises the following steps: collecting the operation efficiency, fault frequency and maintenance record of a machine tool through real-time monitoring data, carrying out the weight distribution of the operation efficiency, fault frequency and maintenance record, and generating weight health data. According to the invention, by analyzing equipment data, including operation efficiency, fault frequency and maintenance records, potential abnormity and fault trends can be found in time. By allocating the weights to the key operation indexes, the generated weighted health data provides quantitative evaluation for the health state of the equipment, so that a maintenance team can perform maintenance in a targeted manner instead of depending on a conventional predetermined maintenance plan. Besides, the health score of the equipment is updated in real time through trend analysis, so that instant fault diagnosis becomes possible, the non-planned downtime of the equipment is reduced, and the overall efficiency of a production line is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of machine tool operation and maintenance, and in particular to a fault monitoring method and system for an intelligent gantry machining center. Background Art

[0002] "CNC machining center", referred to as "CNC machining center" or "CNC center", is an advanced mechanical processing equipment (referred to as machine tool), which uses a computer control system to accurately control multi-axis motion to achieve automated processing of complex parts. In the field of industrial manufacturing, it is a machine tool equipment with advanced technology, high precision, high speed, high degree of automation and multi-functional characteristics.

[0003] In the existing technology, fault monitoring of intelligent gantry machining centers relies on scheduled maintenance and operator experience to judge faults. This maintenance mode that relies on experience rather than data can handle common problems, but it may lead to misjudgment or missed judgment for complex or uncommon faults, increasing safety risks. In addition, once the equipment fails, it may have caused significant damage or production delays, affecting the reliability and operational safety of the equipment. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a fault monitoring method and system for an intelligent gantry machining center.

[0005] In order to achieve the above object, the present invention adopts the following technical solution, a fault monitoring method for an intelligent gantry machining center, comprising the following steps:

[0006] Through real-time monitoring data, the operating efficiency, failure frequency and maintenance records of machine tools are collected, and weights are assigned to the operating efficiency, failure frequency and maintenance records to generate weighted health data;

[0007] Based on the weighted health data, processing each indicator in the weighted health data, calculating and updating the real-time health score of each machine tool, performing trend analysis on the real-time health score, detecting changes in the real-time health score, and generating a health trend result;

[0008] Based on the health trend results, identify abnormal patterns, analyze the differences between normal and abnormal operating data, adjust recognition parameters to optimize recognition accuracy, and generate an adjusted anomaly detection model;

[0009] Based on the adjusted anomaly detection model, real-time monitoring is implemented, abnormal events are identified and recorded, the causes and impacts of abnormal events are analyzed, an abnormal event database is established, maintenance needs faced by the machine tool are analyzed through the abnormal event database, and a maintenance list is generated.

[0010] Preferably, the steps for obtaining the weighted health data are:

[0011] Based on real-time monitoring data, the operating efficiency, failure frequency and maintenance records of machine tools are collected to obtain operating data;

[0012] Calculating weighted scores of operating efficiency, failure frequency and maintenance record respectively according to the operating data;

[0013] Based on the weighted scores, weighted health data of the machine tool is generated.

[0014] Preferably, the steps for obtaining the real-time health score are:

[0015] Based on the weighted health data, extract the index value of each machine tool to obtain index data;

[0016] Based on the indicator data, a real-time health score for each machine tool is calculated.

[0017] Preferably, the steps for obtaining the health trend results are:

[0018] Collect the real-time health score of each machine tool within a specified time period to form a health score time series;

[0019] Performing time series analysis on the health score time series, identifying and quantifying the changing trend of the score, detecting fluctuations in health status, and obtaining analysis results;

[0020] Based on the analysis results, trends and patterns are identified and health trend results for the machine tool are generated.

[0021] Preferably, the steps of obtaining the adjusted anomaly detection model are:

[0022] Based on the health trend results, using machine learning to identify data patterns that deviate from normal operating trends, and marking the data patterns that deviate from normal operating trends as abnormal patterns;

[0023] Compare the operating data under normal and abnormal modes, analyze the differences in operating efficiency, failure frequency and maintenance frequency, and determine the factors that affect the fluctuation of health scores;

[0024] Adjust the machine learning parameters, reapply the adjusted machine learning parameters, run the test data to verify the effect, and generate the adjusted anomaly detection model.

[0025] Preferably, the steps of acquiring the abnormal event database are:

[0026] Deploy the adjusted anomaly detection model, and when the anomaly detection model identifies an abnormal pattern, record the event time, the affected machine tool number, and the characteristics of the abnormal pattern to form an abnormal event record;

[0027] Analyze the abnormal event records, compare the operating data and maintenance records before and after the abnormality occurs, identify the relationship between the abnormal event and the machine tool performance, determine the cause and potential impact of each abnormal event, and generate an abnormal event database.

[0028] Preferably, the steps of obtaining the maintenance list are:

[0029] Based on the data in the abnormal event database, identify the machine tools and components that have failed, analyze the mode and frequency of the failure, determine the machine tools or components that need immediate maintenance or scheduled maintenance, and obtain analysis results;

[0030] A maintenance list is compiled according to the analysis result, wherein the maintenance list includes the maintenance type, maintenance time and required resources for each machine tool.

[0031] The present invention provides a fault monitoring system, comprising:

[0032] The fault data collection module monitors the real-time operation data of the machine tool, collects the operation efficiency, fault frequency and maintenance records of the machine tool, assigns weights to the operation efficiency, fault frequency and maintenance records, and generates weighted health data;

[0033] The health score analysis module processes the indicators in the health data based on the weighted health data, calculates and updates the real-time health score of the machine tool, performs trend analysis on the score, and generates health trend results;

[0034] The abnormality identification and maintenance decision module identifies abnormal patterns based on health trend results, compares the differences between normal mode and abnormal mode operating data, records abnormal events and analyzes the causes, establishes an abnormal event database, analyzes the maintenance needs of machine tools through the abnormal event database, and generates a maintenance recommendation list.

[0035] Compared with the prior art, the advantages and positive effects of the present invention are:

[0036] The present invention can timely detect potential anomalies and failure trends by analyzing equipment data, including operating efficiency, failure frequency, and maintenance records. By assigning weights to key operating indicators, the generated weighted health data provides a quantitative assessment of the equipment health status, allowing the maintenance team to perform targeted maintenance instead of relying on conventional scheduled maintenance plans. In addition, the health score of the equipment is updated in real time through trend analysis, making instant fault diagnosis possible, thereby reducing unplanned downtime of the equipment and improving the overall efficiency of the production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0039] See also Figure 1 The present invention provides a technical solution, a fault monitoring method for an intelligent gantry machining center, comprising the following steps:

[0040] Through real-time monitoring data, the operating efficiency, failure frequency and maintenance records of machine tools are collected, and weights are assigned to the operating efficiency, failure frequency and maintenance records to generate weighted health data;

[0041] Based on the weighted health data, process each indicator in the weighted health data, calculate and update the real-time health score of each machine tool, perform trend analysis on the real-time health score, detect changes in the real-time health score, and generate health trend results;

[0042] Based on the health trend results, identify abnormal patterns, analyze the differences between normal and abnormal operation data, adjust the recognition parameters to optimize the recognition accuracy, and generate an adjusted anomaly detection model;

[0043] Based on the adjusted anomaly detection model, real-time monitoring is implemented to identify and record abnormal events, analyze the causes and impacts of abnormal events, establish an abnormal event database, analyze the maintenance needs faced by machine tools through the abnormal event database, and generate a maintenance list.

[0044] The steps to obtain weighted health data are:

[0045] Based on real-time monitoring data, the operating efficiency, failure frequency and maintenance records of machine tools are collected to obtain operating data;

[0046] According to the operation data, the weighted scores of operation efficiency, failure frequency and maintenance record are calculated respectively. The calculation formula is:

[0047]

[0048] Among them, S is the weighted score, w i is the weight of data item i, x i is the observed value of data item i, and n is the total number of data items;

[0049] Based on the weighted scores, weighted health data of the machine tool is generated.

[0050] Specifically, the operating efficiency, failure frequency and maintenance records of relevant machine tools are collected, and the operating parameters of each machine tool are recorded in detail, such as the operating time per hour, the number of failure shutdowns and the number of operating hours after each maintenance. These data are automatically stored in the database for subsequent analysis and to obtain operating data.

[0051] formula The benefit is that by dynamically adjusting the weight w i , which can accurately reflect the importance of different machine tool performance indicators, thereby achieving a more personalized health assessment; parameter w i is the weight of data item i, which is determined by analyzing the importance of various indicators on the health of machine tool operation and combining historical data and expert experience; parameter x i is the observed value of data item i, which is obtained by directly collecting the operating efficiency, failure frequency and maintenance records of the machine tool.

[0052] Calculation process: Set the weight of the machine tool's operating efficiency to 0.5, the weight of the fault frequency to 0.3, and the weight of the maintenance record to 0.2. If the operating efficiency score of a machine tool is 85 points, the fault frequency score is 20 times, and the number of maintenance times is 3 times, then:

[0053] S=(0.5×85)+(0.3×20)+(0.2×3)=42.5+6+0.6=49.1

[0054] The results show that the overall health score of the machine tool is 49.1. If the score is lower than the set health threshold, it indicates that the machine tool may need maintenance or inspection to ensure operating efficiency.

[0055] Based on the weighted scores, the scores are normalized. Data normalization can be used to standardize the scores of each machine tool to the same comparison level, which is convenient for subsequent comparison and ranking of the health status of the machine tools and generates the final weighted health data as the basis for evaluating the maintenance priority of each machine tool.

[0056] The steps to obtain real-time health scores are:

[0057] Based on the weighted health data, the index value of each machine tool is extracted to obtain the index data;

[0058] According to the indicator data, the real-time health score of each machine tool is calculated using the following formula:

[0059]

[0060] Among them, H is the real-time health score, r j is the importance of indicator j, d j is the observed value of indicator j, m is the total number of indicators, and p is the normalized index.

[0061] Specifically, based on the weighted health data extraction process, the weighted health data includes the operating efficiency, failure frequency and maintenance records of the machine tools. These indicators are used to evaluate the overall operating status of the machine tools. Data is continuously collected through the real-time monitoring system. Each indicator is recorded and time series analysis is performed to determine the performance of each machine tool under different working conditions. This continuous monitoring and recording of data ensures the integrity and real-time nature of the data. Through these comprehensive data, the actual operating status of each machine tool can be more accurately reflected, thereby better conducting health assessments.

[0062] formula The benefit of the parameter r is that the sensitivity of the score is enhanced through normalization and exponential operation, so that the score can more dynamically reflect the actual impact of changes in various indicators of the machine tool on the health status; j is the relative importance of index j, which is determined by statistical analysis of the historical maintenance and failure data of the machine tool; parameter d j is the actual observed value of indicator j, which is directly collected through the real-time monitoring system; parameter m is the total number of indicators, which is determined according to the number of monitoring indicators set; parameter p is the normalized index, which is obtained by performing regression analysis and sensitivity testing on historical data to adjust the response sensitivity of the scoring system.

[0063] Calculation process: The machine tool has three indicators, recorded as d1, d2, and d3, respectively. The corresponding importance weights are r1=0.3, r2=0.5, and r3=0.2. The normalization index p=2. The real-time observation values ​​are d1=50, d2=30, and d3=20. First, the weighted normalized values ​​of each indicator are calculated. Then apply the formula to calculate the score:

[0064]

[0065] The result shows that based on the current indicator values ​​and the set weights, the machine tool has a low health score, indicating that it may need maintenance or further inspection to determine the cause of the fault.

[0066] Based on the update process of the real-time health score, the real-time health score is a dynamically calculated indicator that reflects the immediate operating status of the machine tool. This indicator is crucial for the maintenance team because it can help the team determine which machine tools need priority maintenance and which machine tools are operating well. The updated health score data will be recorded in the maintenance database to guide subsequent maintenance decisions and the formulation of preventive maintenance plans.

[0067] The steps to obtain health trend results are:

[0068] Collect the real-time health score of each machine tool within a specified time period to form a health score time series;

[0069] Perform time series analysis on the health score time series, identify and quantify the changing trend of the score, detect health status fluctuations, and obtain analysis results;

[0070] Based on the analysis results, trends and patterns are identified and health trend results for the machine tool are generated.

[0071] Specifically, during a predetermined monitoring period, the operating parameters of each machine tool, such as current, temperature, vibration, etc., are collected in real time. These parameters are converted into health scores by algorithms, and these scores are continuously recorded over time to form a detailed health score time series, which provides basic data for subsequent analysis;

[0072] From the health score time series, the autoregressive moving average model or seasonal difference autoregressive process is used to analyze the data and process the seasonal and non-seasonal fluctuations in the data. The trend of the health score can be calculated and the possible short-term failure and long-term performance degradation of the machine tool can be detected. This process not only identifies the frequency and amplitude of the score fluctuation, but also evaluates its potential impact on production efficiency.

[0073] Based on the details of the time series analysis, key change points and trend patterns in the health scores are identified, such as sudden increases or decreases. These trends are further classified as temporary or permanent changes. For each pattern, this information is summarized and converted into health trend results. These results describe in detail how the health status of each machine tool evolves over time, providing the maintenance team with a dynamic health map of the operating machine tools to guide future maintenance and operation decisions.

[0074] The steps to obtain the adjusted anomaly detection model are:

[0075] Based on the health trend results, machine learning is used to identify data patterns that deviate from normal operating trends, and the data patterns that deviate from normal operating trends are marked as abnormal patterns;

[0076] Compare the operating data under normal and abnormal modes, analyze the differences in operating efficiency, failure frequency and maintenance frequency, and determine the factors that affect the fluctuation of health scores;

[0077] Adjust the machine learning parameters, reapply the adjusted machine learning parameters, run the test data to verify the effect, and generate the adjusted anomaly detection model.

[0078] Specifically, based on the health trend results, machine learning methods are used to analyze the data, and random forests or support vector machines are used to detect abnormal patterns in the data. The data differences between normal operation and fault conditions of various machine tools are identified through training models, so as to mark data patterns that deviate from normal operating trends and mark data patterns that deviate from normal operating trends as abnormal patterns.

[0079] A comparative analysis is performed on the marked abnormal and normal mode data sets. By calculating and comparing the operating efficiency, failure frequency, and maintenance frequency in the two modes, the significant differences in these parameters in different modes are identified, thereby clarifying the behavior patterns under different machine tool states and determining the key factors affecting the fluctuation of the machine tool health score. These factors provide a scientific basis for subsequent model adjustments and determine the factors affecting the fluctuation of the health score.

[0080] Based on the analysis results of the first two steps, the parameters of the machine learning model are adjusted, such as adjusting the classification threshold, learning rate or complexity parameters to improve the model's adaptability to new situations and prediction accuracy. The adjusted parameters are reapplied and the model is run on a new test dataset to verify the effect. This step involves iterative optimization and model retraining to ensure that the model update can reflect the latest data features and abnormal behaviors, and finally generate an adjusted anomaly detection model, which provides higher accuracy and reliability in practical applications.

[0081] The steps for obtaining the abnormal event database are as follows:

[0082] Deploy the adjusted anomaly detection model. When the anomaly detection model identifies an abnormal pattern, it records the event time, the affected machine tool number, and the characteristics of the abnormal pattern to form an abnormal event record.

[0083] Analyze abnormal event records, compare operating data and maintenance records before and after the abnormality occurs, identify the relationship between abnormal events and machine tool performance, determine the cause and potential impact of each abnormal event, and generate an abnormal event database.

[0084] Specifically, once the adjusted anomaly detection model is deployed and runs on the production line, the model begins to monitor the data output of the machine tool in real time and evaluates each reading of the machine tool status. The model uses built-in recognition rules to analyze data points. When the data shows fluctuations consistent with the established abnormal pattern, an alarm is triggered. These alarm activities include recording the exact time when the anomaly occurred, associating the machine tool number at this time, and recording the characteristics of the abnormal pattern in detail. This information is automatically archived as an abnormal event record, providing the necessary information and data support for subsequent fault diagnosis and processing, thereby forming an initial response and recording process for abnormal situations.

[0085] For abnormal event records that have been formed, the operation data and maintenance records related to the records are extracted by directly accessing the machine tool's operation database. Next, these data are compared to determine the change points before and after the abnormality occurs, especially the changes in failure frequency and maintenance requirements. Through this basic data comparison, the environment and conditions under which the abnormality occurs can be portrayed. Further, by checking each record one by one, the specific factors that lead to performance degradation and how these factors affect the operating efficiency of the machine tool can be analyzed and determined.

[0086] The steps to obtain the maintenance list are:

[0087] Based on the data in the abnormal event database, identify the machine tools and components that have failed, analyze the mode and frequency of failure, determine the machine tools or components that need immediate maintenance or scheduled maintenance, and obtain analysis results;

[0088] Based on the analysis results, a maintenance list is compiled, which includes the maintenance type, maintenance time and required resources for each machine tool.

[0089] Specifically, extract relevant data from the abnormal event database to identify the machine tools and components with frequent failures, which involves classification and frequency analysis of failure events, using clustering algorithms to identify patterns and failure trends, and calculating the frequency of each type of failure in order to identify the machine tools and components with frequent problems. This analysis is based on the type of failure, frequency of occurrence, affected machine tool components, and operating conditions when the failure occurs. Through these data analyses, the machine tools or components that require immediate or regular maintenance due to repeated failures are determined, thereby providing a basis for the next maintenance plan and forming an analysis result that includes the machine tools and components that need maintenance;

[0090] Based on the above analysis results, a maintenance plan is prepared. The plan lists in detail the specific maintenance types for each machine tool that needs maintenance, including preventive maintenance, emergency repairs, etc., and specifies the specific maintenance time and the resources required to perform maintenance. These resources include necessary tools, replacement parts, technicians and their skill levels. The preparation of the maintenance plan takes into account the machine tool's operating history, fault type, frequency and its impact to ensure that each machine tool is properly maintained according to its operating status and fault history, thereby reducing future operating interruptions and generating a maintenance list that provides a clear execution blueprint for future maintenance operations.

[0091] The present invention provides a fault monitoring system, comprising:

[0092] The fault data collection module monitors the real-time operation data of the machine tool, collects the operation efficiency, fault frequency and maintenance records of the machine tool, assigns weights to the operation efficiency, fault frequency and maintenance records, and generates weighted health data;

[0093] The health score analysis module processes the indicators in the health data based on the weighted health data, calculates and updates the real-time health score of the machine tool, performs trend analysis on the score, and generates health trend results;

[0094] The abnormality identification and maintenance decision module identifies abnormal patterns based on health trend results, compares the differences between normal mode and abnormal mode operating data, records abnormal events and analyzes the causes, establishes an abnormal event database, analyzes the maintenance needs of machine tools through the abnormal event database, and generates a maintenance recommendation list.

[0095] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A fault monitoring method for an intelligent gantry machining center, characterized in that: The following steps are involved: Through real-time monitoring data, the operating efficiency, failure frequency and maintenance records of machine tools are collected, and weights are assigned to the operating efficiency, failure frequency and maintenance records to generate weighted health data; Based on the weighted health data, processing each indicator in the weighted health data, calculating and updating the real-time health score of each machine tool, performing trend analysis on the real-time health score, detecting changes in the real-time health score, and generating a health trend result; Based on the health trend results, identify abnormal patterns, analyze the differences between normal and abnormal operating data, adjust recognition parameters to optimize recognition accuracy, and generate an adjusted anomaly detection model; Based on the adjusted anomaly detection model, real-time monitoring is implemented, abnormal events are identified and recorded, the causes and impacts of abnormal events are analyzed, an abnormal event database is established, maintenance needs faced by the machine tool are analyzed through the abnormal event database, and a maintenance list is generated.

2. The fault monitoring method of the intelligent gantry machining center according to claim 1 is characterized in that: The steps for obtaining the weighted health data are as follows: Based on real-time monitoring data, the operating efficiency, failure frequency and maintenance records of machine tools are collected to obtain operating data; Calculating weighted scores of operating efficiency, failure frequency and maintenance record respectively according to the operating data; Based on the weighted scores, weighted health data of the machine tool is generated.

3. The fault monitoring method of the intelligent gantry machining center according to claim 1 is characterized in that: The steps for obtaining the real-time health score are as follows: Based on the weighted health data, extract the index value of each machine tool to obtain index data; Based on the indicator data, a real-time health score for each machine tool is calculated.

4. The fault monitoring method of the intelligent gantry machining center according to claim 1 is characterized in that: The steps for obtaining the health trend results are: Collect the real-time health score of each machine tool within a specified time period to form a health score time series; Performing time series analysis on the health score time series, identifying and quantifying the changing trend of the score, detecting fluctuations in health status, and obtaining analysis results; Based on the analysis results, trends and patterns are identified and health trend results for the machine tool are generated.

5. The fault monitoring method of the intelligent gantry machining center according to claim 1 is characterized in that: The steps for obtaining the adjusted anomaly detection model are: Based on the health trend results, using machine learning to identify data patterns that deviate from normal operating trends, and marking the data patterns that deviate from normal operating trends as abnormal patterns; Compare the operating data under normal and abnormal modes, analyze the differences in operating efficiency, failure frequency and maintenance frequency, and determine the factors that affect the fluctuation of health scores; Adjust the machine learning parameters, reapply the adjusted machine learning parameters, run the test data to verify the effect, and generate the adjusted anomaly detection model.

6. The fault monitoring method of the intelligent gantry machining center according to claim 1, characterized in that: The steps for acquiring the abnormal event database are: Deploy the adjusted anomaly detection model, and when the anomaly detection model identifies an abnormal pattern, record the event time, the affected machine tool number, and the characteristics of the abnormal pattern to form an abnormal event record; Analyze the abnormal event records, compare the operating data and maintenance records before and after the abnormality occurs, identify the relationship between the abnormal event and the machine tool performance, determine the cause and potential impact of each abnormal event, and generate an abnormal event database.

7. The fault monitoring method of the intelligent gantry machining center according to claim 1, characterized in that: The steps for obtaining the maintenance list are: Based on the data in the abnormal event database, identify the machine tools and components that have failed, analyze the mode and frequency of the failure, determine the machine tools or components that need immediate maintenance or scheduled maintenance, and obtain analysis results; A maintenance list is compiled according to the analysis result, wherein the maintenance list includes the maintenance type, maintenance time and required resources for each machine tool.

8. A fault monitoring system for an intelligent gantry machining center fault monitoring method according to any one of claims 1 to 7, characterized in that: include: The fault data collection module monitors the real-time operation data of the machine tool, collects the operation efficiency, fault frequency and maintenance records of the machine tool, assigns weights to the operation efficiency, fault frequency and maintenance records, and generates weighted health data; The health score analysis module processes the indicators in the health data based on the weighted health data, calculates and updates the real-time health score of the machine tool, performs trend analysis on the score, and generates health trend results; The abnormality identification and maintenance decision module identifies abnormal patterns based on health trend results, compares the differences between normal mode and abnormal mode operating data, records abnormal events and analyzes the causes, establishes an abnormal event database, analyzes the maintenance needs of machine tools through the abnormal event database, and generates a maintenance recommendation list.

Images