Numerical control machining center reliability modeling and risk analysis method

Abstract

The invention discloses a reliability modeling and risk analysis method for a numerical control machining center. The method comprises the following steps: acquiring field operation and maintenance data of the numerical control machining center and dividing the whole machine into a plurality of functional subsystems; constructing a hybrid Weibull distribution model for each subsystem, determining an optimal hybrid component number by using AIC and BIC criteria, and performing parameter estimation by using an EM algorithm to characterize life characteristics hybrid with multiple failure mechanisms; based on a competitive risk theory, calculating a cumulative fault probability CFP of each subsystem and a first fault contribution degree FFC to complete machine failure; and constructing a risk quadrant graph by taking the fault frequency and the first fault contribution degree as dimensions, evaluating the risk region category of each subsystem, and generating a differentiated maintenance strategy. According to the method, the problem that a single distribution model is difficult to adapt to heterogeneous fault data is solved, and the actual influence of the subsystem on the shutdown of the whole machine can be quantitatively evaluated, so that a scientific preventive maintenance plan is guided to be formulated.

Description

Technical Field

[0001] This invention belongs to the field of CNC machine tool reliability engineering technology, and relates to a method for reliability modeling and risk analysis of CNC machining centers. Specifically, it relates to a method for reliability modeling, competitive risk assessment and maintenance decision generation of CNC machining center subsystems that considers the heterogeneity of fault data. Background Technology

[0002] CNC machining centers are core equipment in modern intelligent manufacturing, widely used in aerospace, automotive manufacturing, precision machinery, and other fields. A CNC machining center is a highly integrated system of multiple functional subsystems, with a complex structure comprising hydraulic, pneumatic, electrical, and spindle subsystems. In actual service, under alternating loads and complex operating conditions, the failure modes of each subsystem exhibit a complex mix of mechanisms, such as random failures caused by early installation defects and aging failures caused by later wear.

[0003] Existing reliability modeling methods for machining centers typically assume that the system or component follows a single lifetime distribution. However, field failure data often exhibits multi-peak or long-tail characteristics, making it difficult for single-distribution models to accurately describe the complex lifetime patterns caused by a mixture of multiple failure mechanisms. This results in low model fitting accuracy and an inability to accurately distinguish the reliability characteristics of equipment at different stages of its life cycle. Although some research has attempted to introduce hybrid distribution models, in engineering practice, there is a lack of a systematic, data-driven method to determine the number of components in the hybrid model. The modeling process often relies on manual experience, leading to strong subjectivity and poor generalization ability.

[0004] Furthermore, in overall system risk assessment, existing analytical methods typically treat each subsystem as an independent risk source, primarily ranking them based on static indicators such as failure frequency or mean time between failures (MTBF). This approach ignores the competing failure relationships between subsystems within the overall system. In a series system, the impact of different subsystem failures on overall system downtime is competitive; failure frequency alone cannot accurately reflect the probability that a subsystem will cause the system to fail first. Some subsystems, although having low failure frequencies, are often the primary cause of overall system downtime when they occur. Traditional assessment methods tend to mask this critical risk, leading to a mismatch between maintenance resource allocation and actual risk requirements.

[0005] Therefore, there is an urgent need for a reliability analysis method that can take into account the heterogeneity of fault data, automatically determine the order of the hybrid model, and combine competing failure theory to quantitatively evaluate the contribution of each subsystem to the overall machine failure, so as to support the accurate maintenance decision-making of CNC machining centers. Summary of the Invention

[0006] The purpose of this invention is to provide a reliability modeling and risk analysis method for CNC machining centers based on a hybrid Weibull distribution. This invention achieves accurate description of heterogeneous fault data by constructing a data-driven hybrid Weibull distribution model; it introduces competing risk theory and a first-failure contribution index to quantitatively evaluate the dynamic contribution of each subsystem to the overall machine failure; and it constructs a fault risk quadrant diagram to enable visualized decision-making for maintenance strategies, thereby guiding machine tool manufacturers and users to formulate scientific and differentiated preventive maintenance plans and improve the overall machine reliability.

[0007] To achieve the above objectives, the technical means employed in this invention are as follows:

[0008] A reliability modeling and risk analysis method for CNC machining centers based on a hybrid Weibull distribution includes the following steps:

[0009] Step S1: Data Acquisition and Subsystem Division. Acquire on-site fault operation and maintenance data of the CNC machining center, calculate the time between failures (TBF), and divide the entire machine into subsystems. Each functional subsystem has its own fault sample set. ;

[0010] Step S2: Construct a hybrid Weibull distribution model, building a model with specific characteristics for each subsystem's fault data. A mixed Weibull distribution model of multiple components is used to characterize the lifetime characteristics of multiple failure mechanisms.

[0011] Step S3: Model order determination and parameter estimation. Based on the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC), determine the optimal number of mixture components for each subsystem. The Expectation-Maximization (EM) algorithm is used for parameter estimation to obtain the shape parameters, scale parameters, and mixing weights of each subsystem. ;

[0012] Step S4: Competition risk and first failure contribution analysis. Based on the competition risk theory, each subsystem is regarded as a competition risk source that leads to the failure of the whole machine. The cumulative failure probability function (CFP) of each subsystem is calculated, and the first failure contribution (FFC) of each subsystem is further calculated.

[0013] Step S5: Risk Quadrant Map Construction and Maintenance Strategy Generation. Using the failure frequency of the machining center subsystem as the horizontal axis and the contribution of the first failure as the vertical axis, a fault risk quadrant map is constructed to divide each subsystem into different risk areas and output differentiated maintenance decision schemes accordingly.

[0014] Furthermore, the probability density function of the mixed Weibull distribution model described in step S2 It is by The composition of Weibull components:

[0015] ……………………………(1)

[0016] In the formula, For time variables, The mixed component of system j's fault data reflects the complexity of the system's lifetime composition. This represents the weight of component h. ; For the scaling parameters of the Weibull model, For shape parameters;

[0017] Then the probability density function is given. With reliability function They are respectively:

[0018] ……………………………(2)

[0019] ……………………………………(3)

[0020] Furthermore, step S3, which uses the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) to determine the optimal number of mixed components for each functional subsystem, includes:

[0021] For each candidate number of components, parameter estimation is performed and the corresponding AIC and BIC values ​​are calculated; the number of components that minimizes the AIC and BIC values ​​is selected as the optimal number of mixed components for the functional subsystem; if the optimal model corresponds to Then the functional subsystem is suitable for a single Weibull distribution; if This indicates that the functional subsystem exhibits significant heterogeneous lifetime characteristics.

[0022] Furthermore, in step S3, the Expectation-Maximization (EM) algorithm is used to estimate the model parameters under the optimal number of mixture components, including iteratively executing the following steps:

[0023] Step E: Based on the current parameter estimates, calculate which step the observed data belongs to. The posterior probability of each mixture component;

[0024] M-step: Update the estimates of the mixed weights, shape parameters, and scale parameters by maximizing the expectation of the log-likelihood function;

[0025] Repeat the E-step and M-step until the log-likelihood function converges, and output the final model parameters.

[0026] Furthermore, the calculation method for the contribution of the first fault in step S4 includes:

[0027] First, calculate the cumulative failure probability function of the functional subsystem. :

[0028] …………………(8)

[0029] In the formula, It is a functional subsystem The fault density function; For the functional subsystem Reliability functions for other subsystems besides those mentioned above;

[0030] Then calculate the functional subsystem. First-fault contribution:

[0031] ……………………………(9)

[0032] In the formula, The set time for the entire machine's task. The total number of subsystems;

[0033] Furthermore, step S5, which involves constructing a fault risk quadrant diagram and generating a maintenance strategy, includes:

[0034] Establish a coordinate system with the failure frequency of each functional subsystem as the horizontal axis and the contribution of the first failure as the vertical axis; calculate the average failure frequency of all functional subsystems. Compared with average contribution and with Divide the coordinate plane into four quadrants with the origin as the origin; divide each subsystem according to its coordinates. Mapping to the corresponding quadrants creates a fault risk quadrant diagram. The four quadrants are defined as follows: Quadrant 1: High frequency, high contribution; Quadrant 2: Low frequency, high contribution; Quadrant 3: High frequency, low contribution; Quadrant 4: Low frequency, low contribution. The functional subsystems in Quadrant 1 perform preventative replacement and key monitoring; the systems in Quadrant 2 perform status monitoring and design improvements; the functional subsystems in Quadrant 3 perform maintenance process optimization and spare parts inventory management; and the systems in Quadrant 4 perform routine inspections or post-incident maintenance.

[0035] A reliability assessment and risk analysis system for CNC machining centers, comprising:

[0036] The data acquisition module is used to acquire fault data of CNC machining centers and divide them into functional subsystems;

[0037] The model building module is used to build a hybrid Weibull distribution model that takes into account data heterogeneity.

[0038] The computational processing module is used to obtain the shape parameters, scale parameters, and mixed weights of each functional subsystem; and to calculate the contribution of the first failure.

[0039] The decision output module is used to generate fault risk quadrant diagrams and maintenance strategy instructions.

[0040] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the reliability modeling and risk analysis method for CNC machining centers as described above.

[0041] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the reliability modeling and risk analysis method for CNC machining centers as described above.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] The reliability modeling and risk analysis method for CNC machining centers of this invention employs a hybrid Weibull distribution to characterize multi-mechanism failure behavior, simultaneously capturing early failure, random failure, and wear-out failure characteristics, with significantly better fitting accuracy than single distribution models. Based on the AIC / BIC criterion, the mixture component is determined using data-driven methods, avoiding subjective arbitrariness and enhancing model interpretability and engineering applicability. By integrating competitive risk analysis and heterogeneous data modeling, a quantitative assessment of the contribution of subsystems to the first failure of the entire machine is achieved, leading to more accurate identification of key failure sources. A fault risk quadrant diagram enables visualized risk classification, supporting dynamic adjustment of maintenance priorities and facilitating optimal overall machine reliability under limited resources. The technology proposed in this invention can be extended to reliability assessment and health management of complex multi-subsystem equipment such as industrial robots and wind power equipment. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a flowchart of the reliability modeling and risk analysis method for CNC machine tool machining centers described in this invention;

[0046] Figure 2 This is a flowchart of the EM parameter estimation process using a hybrid Weibull distribution in this invention.

[0047] Figure 3 This is the risk quadrant diagram based on fault frequency and fault contribution in this invention;

[0048] Figure 4a This is a comparison chart of the reliability curves of the single Weibull model and the hybrid Weibull model of the cooling system in this invention;

[0049] Figure 4b This is a comparison chart of the reliability curves of the single Weibull model and the hybrid Weibull model of the lubrication system in this invention. Detailed Implementation

[0050] The present invention will now be described in detail with reference to the accompanying drawings:

[0051] See Figure 1 This invention provides a reliability modeling and risk analysis method for CNC machining centers based on hybrid Weibull distribution, comprising the following five stages: data acquisition and subsystem partitioning, hybrid Weibull modeling construction, model order determination and parameter estimation, competitive risk and first failure contribution analysis, risk quadrant diagram construction and maintenance strategy generation.

[0052] The specific methods for fault data acquisition and subsystem division in step S1 are as follows:

[0053] (1) Collect on-site fault data of the target CNC machining center within the set tracking period. Each record shall include at least the equipment number, fault occurrence time, fault location, fault phenomenon, maintenance start time and maintenance end time.

[0054] (2) For the same machining center, calculate the fault interval (TBF) for each fault based on the fault occurrence time and the end time of the previous maintenance cycle.

[0055] (3) Based on the structure and functional principles of the CNC machining center, the whole machine is divided into multiple subsystems, and each fault record is mapped to the corresponding subsystem;

[0056] (4) For each subsystem, organize all its corresponding TBF data to form the lifetime dataset of that subsystem;

[0057] The specific method for constructing the hybrid Weibull distribution model in step S2 is as follows:

[0058] For the lifetime dataset of each subsystem, construct a system containing... A mixture distribution model of Weibull components, with its probability density function as follows:

[0059] ……………………………(1)

[0060] In the formula, For time variables, The number of components in the mixture. Let h be the weight of the h-th component, and satisfy the following condition: , For scale parameters, For shape parameters;

[0061] Then the probability density function is given. With reliability function They are respectively:

[0062] ………………………(2)

[0063] …………………………(3)

[0064] Based on the concept of multi-mechanism hybridity The fault data of subsystem j is heterogeneous, with each mixed component corresponding to a lifetime group and representing a type of failure.

[0065] The specific methods for model order determination and parameter estimation in step S3 are as follows:

[0066] (1) Model selection, setting the candidate mixture component set For each candidate Calculate its AIC and BIC values:

[0067] …………………………………………(4)

[0068] ………………………………………(5)

[0069] in, The number of model parameters, For the sample size, For log-likelihood, the model that minimizes both AIC and BIC values ​​is selected as the optimal model for the subsystem. If the optimal model corresponds to... If the subsystem is subject to a single Weibull distribution; This indicates that the subsystem exhibits significant heterogeneous lifetime characteristics;

[0070] (2) Parameter estimation: The expectation-maximization (EM) algorithm is used to iteratively estimate the model parameters. See [reference needed]. Figure 2 The algorithm converges by iteratively performing the following two steps:

[0071] 1) Step E: Calculate each sample The posterior probability of belonging to component h :

[0072] ……………(6)

[0073] Indicates the first Fault interval time corresponding to each sample;

[0074] Indicates the number of iterations;

[0075] Indicates the first The first subsystem The scale parameters corresponding to each component;

[0076] Indicates the first The first subsystem The shape parameters corresponding to each component.

[0077] 2) M-step: based on Update weights For each component h, in the weighted sample Maximize the weighted log-likelihood:

[0078] ………………(7)

[0079] Indicates the first Failure frequency of each subsystem; Represents a sample set;

[0080] The specific method for analyzing the competition risk and first failure contribution in step S4 is as follows:

[0081] (1) Establish a competitive risk model at the whole machine level, and treat each subsystem as an independent potential cause of failure;

[0082] (2) Based on the optimal hybrid Weibull model of each subsystem, calculate its reliability function respectively. With fault distribution function ;

[0083] (3) Calculate the cumulative failure probability function (CFP) of subsystem j in time t, which represents the probability that the subsystem will be the first to cause the failure of the whole machine in time t. The calculation formula is as follows:

[0084] …………(8)

[0085] In the formula, It is a functional subsystem The fault density function; For the functional subsystem Reliability functions for other subsystems besides those mentioned above;

[0086] y means except Other subsystems besides;

[0087] It is equivalent to the "intermediate time point" swept from 0 to t in d(t) (used to run through the time in the integration). It has no additional independent meaning; it is just a time parameter that needs to be introduced for integration.

[0088] (4) Set a representative whole machine task time. Calculate the cumulative failure probability value of each subsystem within this time period. The values ​​are normalized to obtain the first-fault contribution of each subsystem. :

[0089] ……………………………(9)

[0090] In the formula, The set time for the entire machine's task. The total number of subsystems; Quantitatively reflects the subsystem The relative importance of the first failure occurring during the entire machine's life cycle;

[0091] The specific steps for generating the risk quadrant chart construction and maintenance strategy in S5 are as follows:

[0092] (1) Statistical analysis of the frequency of failures in each subsystem during the observation period. A two-dimensional coordinate system is constructed with fault frequency as the horizontal axis and the contribution of the first fault as the vertical axis.

[0093] (2) Calculate the average failure frequency of all subsystems within any time t. Compared with average contribution and with Divide the coordinate system into four quadrants with the origin as the origin;

[0094] (3) Each subsystem is divided according to its coordinates Mapping to the corresponding quadrants forms a fault risk quadrant map. The four quadrants are defined as follows: Quadrant 1: High frequency and high contribution area; Quadrant 2: Low frequency and high contribution area; Quadrant 3: High frequency and low contribution area; Quadrant 4: Low frequency and low contribution area.

[0095] (4) Based on the risk area to which the subsystem belongs, formulate differentiated maintenance strategy recommendations: implement preventive replacement and key monitoring for high-frequency, high-contribution subsystems; strengthen condition detection and design improvement for low-frequency, high-contribution subsystems; optimize maintenance process and spare parts reserve for high-frequency, low-contribution subsystems; and perform routine monitoring for low-frequency, low-contribution subsystems.

[0096] This completes the reliability modeling and risk analysis of the machining center based on the hybrid Weibull distribution.

[0097] A reliability assessment and risk analysis system for CNC machining centers, comprising:

[0098] The data acquisition module is used to acquire fault data of CNC machining centers and divide them into functional subsystems;

[0099] The model building module is used to build a hybrid Weibull distribution model that takes into account data heterogeneity.

[0100] The computational processing module is used to obtain the shape parameters, scale parameters, and mixed weights of each functional subsystem; and to calculate the contribution of the first failure.

[0101] The decision output module is used to generate fault risk quadrant diagrams and maintenance strategy instructions.

[0102] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the aforementioned method for reliability modeling and risk analysis of CNC machining centers.

[0103] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned method for reliability modeling and risk analysis of CNC machining centers.

[0104] Example

[0105] The following specific engineering embodiment further illustrates the implementation process and effects of the present invention.

[0106] This invention uses a certain model of vertical CNC machining center as an example to illustrate the implementation. A total of 108 fault data points were collected from 36 of these machining centers. The entire machine was divided into 11 subsystems (as shown in Table 1). After mapping, the distribution of the number of samples in each subsystem is shown in Table 2.

[0107] Table 1. Subsystem Division of CNC Machining Center

[0108]

[0109] Table 2 Statistical characteristics of the time between failures (TBF) for each subsystem

[0110]

[0111] Taking the spindle, electrical, and lubrication systems as examples, the TBF data of the subsystems are calculated according to the model order determination method described in step S3. Combined with formulas (4) and (5), the judgment results are obtained by judging according to the AIC / BIC criteria, as shown in Table 3.

[0112] Table 3. Comparison of AIC / BIC of representative subsystem hybrid Weibull models

[0113]

[0114] According to the method for constructing the mixed Weibull distribution model in step S2, i.e., equations (1)-(3), and combined with the parameter estimation method in step S3, after determining the number of mixed components in each subsystem, the EM algorithm is applied to estimate the parameters of the mixed Weibull model of each subsystem according to equations (6) and (7). The estimation results are shown in Table 4.

[0115] Table 4. Parameter estimation results of the hybrid Weibull model for the machining center subsystem

[0116]

[0117] Based on the Weibull parameter estimation results of each subsystem in Table 4, it can be seen that the cooling system and lubrication system adopted a two-component mixed Weibull model, while the other nine subsystems adopted a single Weibull model. The cooling system (3) and lubrication system (9) showed obvious mixed characteristics. Among them, the lifespan of one group in the cooling system was relatively short and the lifespan of the other group was relatively long, but the weight of both groups was 0.5, indicating that the short-life and long-life failure groups within it were relatively balanced. The weights of the two groups in the lubrication system were somewhat different. One group had higher scale and shape parameters, with a weight of about 0.71, while the other group had smaller scale parameters and larger shape parameters, with a weight of about 0.29, indicating that the proportion of high-quality lubrication groups was higher, while the proportion of early failure groups caused by poor lubrication was relatively smaller.

[0118] Based on the parameter estimation results of equations (8) and (9) in step S4 and Table 4, the cumulative failure probability and the contribution of the first failure of each subsystem are calculated. For ease of quantitative comparison, the following parameters are selected: h (twice the maximum TBF of the sample) was calculated, and the results are shown in Table 5.

[0119] Table 5 Cumulative failure probability and first failure contribution of machining center subsystem ( )

[0120]

[0121] Based on the risk visualization and decision support method described in step S5, a risk quadrant diagram is drawn based on failure frequency and contribution (e.g., ...). Figure 3 As shown). Among them:

[0122] (1) First quadrant: High frequency-high contribution area, including hydraulic system and cooling system. The hydraulic system has a high failure frequency and the highest failure contribution of 23.89%. The cooling system is on the edge of this area with a failure contribution of 10.35%. It is recommended to focus on prevention and maintenance of the subsystems in this quadrant.

[0123] (2) Second quadrant: Low frequency - high contribution area, including pneumatic system, workbench and electrical system. These subsystems do not fail frequently, but their contribution to the failure of the whole machine is relatively high. In other words, although these subsystems are not easy to fail, once they fail, the impact on the whole machine is significant. It is recommended to strengthen the monitoring and design optimization of these subsystems.

[0124] (3) Third quadrant: High frequency - low contribution area, including spindle system and tool magazine, among which tool magazine has the highest failure frequency (19 times). The failure frequency of this type of subsystem is high, but it is usually a non-downtime failure that can be quickly recovered. Although it has little impact on the failure of the whole machine, the maintenance cost accounts for a high proportion. It is recommended to optimize the maintenance response and spare parts strategy.

[0125] (4) Fourth quadrant: Low frequency - low contribution area, including CNC system, lubrication system, auxiliary system and feed system. These subsystems have a relatively small impact on the reliability of the machining center and can be managed in a routine manner.

[0126] To verify the superiority of the method of this invention, the fitting accuracy of the hybrid Weibull model is compared with that of the traditional single Weibull model. Taking cooling and lubrication systems as examples, the reliability curves of the single Weibull and hybrid Weibull models are compared, as shown in the figure. Figure 4a , Figure 4b As shown in Table 6, the error analysis results are as follows.

[0127] Table 6 Error Analysis Results

[0128]

[0129] The results show that the fitting error of the hybrid Weibull model is significantly reduced, proving that it can more accurately describe multi-mechanism failure behavior and provide a more scientific basis for reliability assessment and maintenance decisions.

[0130] In summary, this embodiment fully demonstrates the entire implementation process of this invention, from data acquisition, heterogeneous modeling, risk quantification to decision visualization. The results show that, in terms of modeling accuracy, compared to traditional single-distribution models, the hybrid Weibull distribution model proposed in this invention can accurately decouple and characterize the complex failure behavior of multiple coexisting mechanisms within the CNC machining center subsystem, significantly reducing fitting errors and improving the confidence level of reliability assessment. In terms of decision support, this invention, through the calculation of the contribution of the first failure and the construction of a risk quadrant diagram, realizes a risk perspective shift from failure frequency to failure impact, providing a quantifiable and more intuitive scientific basis for the formulation of differentiated maintenance strategies. This effectively solves the problem of coexistence of "over-maintenance" and "under-maintenance" in traditional maintenance, helping enterprises maximize both overall machine reliability and economic benefits with limited maintenance resources.

[0131] Furthermore, the method framework proposed in this invention has high versatility and scalability. It is not only applicable to CNC machining centers, but can also be widely applied to other complex high-end equipment with multi-subsystem coupling and multiple failure mechanisms, such as industrial robots, wind turbine generators, and aero engines. This provides strong technical support for equipment failure prediction and health management in the context of intelligent manufacturing and has broad engineering application prospects.

Claims

1. A method for reliability modeling and risk analysis of CNC machining centers, characterized in that, Includes the following steps: Step S1: Obtain the on-site operation and maintenance data of the CNC machining center, calculate the fault interval time, and divide the entire CNC machining center into several functional subsystems; Step S2: For each of the functional subsystems, construct a hybrid Weibull distribution model, which contains several Weibull components to characterize the lifetime characteristics of different failure mechanisms mixed within the functional subsystem. Step S3: Determine the optimal number of mixing components for each functional subsystem using the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC), and estimate the model parameters under the optimal number of mixing components using the Expectation Maximization (EM) algorithm to obtain the shape parameters, scale parameters, and mixing weights of each functional subsystem. Step S4: Based on the theory of competitive risk, each functional subsystem is regarded as a competitive risk source that leads to the failure of the whole machine. The cumulative failure probability function of each functional subsystem is calculated, and the contribution of each functional subsystem to the first failure of the whole machine is calculated based on the function. Step S5: Construct a fault risk quadrant and generate maintenance strategies. Using fault frequency as the horizontal axis and the contribution of the first fault as the vertical axis, construct a fault risk quadrant, classify risk categories, and output differentiated maintenance decision recommendations.

2. The reliability modeling and risk analysis method for CNC machining centers according to claim 1, characterized in that: The probability density function of the mixed Weibull distribution model described in step S2 It is by The composition of Weibull components: ……………………………（1） In the formula, For time variables, The mixed component of system j's fault data reflects the complexity of the system's lifetime composition. This represents the weight of component h. ; For the scaling parameters of the Weibull model, For shape parameters; Then the probability density function is given. With reliability function They are respectively: ……………………………（2） ……………………………………（3）。 3. The reliability modeling and risk analysis method for CNC machining centers according to claim 2, characterized in that: Step S3, which uses the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) to determine the optimal number of mixed components for each functional subsystem, includes: For each candidate number of components, parameter estimation is performed and the corresponding AIC and BIC values ​​are calculated; the number of components that minimizes the AIC and BIC values ​​is selected as the optimal number of mixed components for the functional subsystem; if the optimal model corresponds to Then the functional subsystem is suitable for a single Weibull distribution; if This indicates that the functional subsystem exhibits significant heterogeneous lifetime characteristics.

4. The reliability modeling and risk analysis method for CNC machining centers according to claim 2, characterized in that: Step S3 uses the Expectation-Maximization (EM) algorithm to estimate the model parameters under the optimal mixture fraction, including iteratively executing the following steps: Step E: Based on the current parameter estimates, calculate which step the observed data belongs to. The posterior probability of each mixture component; M-step: Update the estimates of the mixed weights, shape parameters, and scale parameters by maximizing the expectation of the log-likelihood function; Repeat the E-step and M-step until the log-likelihood function converges, and output the final model parameters.

5. The reliability modeling and risk analysis method for CNC machining centers according to claim 1, characterized in that: The calculation method for the contribution of the first failure in step S4 includes: First, calculate the cumulative failure probability function of the functional subsystem. : …………………（8） In the formula, It is a functional subsystem The fault density function; For the functional subsystem Reliability functions for other subsystems besides those mentioned above; Then calculate the functional subsystem. First-fault contribution: ……………………………（9） In the formula, The set time for the entire machine's task. This represents the total number of subsystems.

6. The reliability modeling and risk analysis method for CNC machining centers according to claim 1, characterized in that: Step S5, which involves constructing a fault risk quadrant diagram and generating a maintenance strategy, includes: Establish a coordinate system with the failure frequency of each functional subsystem as the horizontal axis and the contribution of the first failure as the vertical axis; calculate the average failure frequency of all functional subsystems. Compared with average contribution and with Divide the coordinate plane into four quadrants with the origin as the origin; divide each subsystem according to its coordinates. Mapping to the corresponding quadrants creates a fault risk quadrant diagram. The four quadrants are defined as follows: Quadrant 1: High frequency, high contribution; Quadrant 2: Low frequency, high contribution; Quadrant 3: High frequency, low contribution; Quadrant 4: Low frequency, low contribution. The functional subsystems in Quadrant 1 perform preventative replacement and key monitoring; the systems in Quadrant 2 perform status monitoring and design improvements; the functional subsystems in Quadrant 3 perform maintenance process optimization and spare parts inventory management; and the systems in Quadrant 4 perform routine inspections or post-incident maintenance.

7. A reliability assessment and risk analysis system for CNC machining centers, characterized in that, include: The data acquisition module is used to acquire fault data of CNC machining centers and divide them into functional subsystems; The model building module is used to build a hybrid Weibull distribution model that takes into account data heterogeneity. The computational processing module is used to obtain the shape parameters, scale parameters, and mixing weights of each functional subsystem; Calculate the contribution of the first failure; The decision output module is used to generate fault risk quadrant diagrams and maintenance strategy instructions.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 6.

Images