A high-end complex equipment reliability analysis method and system

Abstract

This invention discloses a method and system for reliability analysis of high-end complex equipment. It establishes correlation strength data and risk factor weight vectors among various complex equipment failures based on typical failure mode data; constructs a risk assessment matrix based on typical failure-related data and correlation strength data between failures; and calculates the risk-benefit value of each failure mode based on the risk factor weight vector and risk assessment matrix. This reduces human factors and uncertainty in the reliability analysis process, effectively improving the objectivity and accuracy of reliability analysis for complex equipment. Based on the failure autocorrelation matrix, this invention incorporates the correlation between different failures and proposes a method for measuring the correlation degree, solving the problem that traditional FMEA does not consider the correlation between different failure modes.

Description

Technical Field

[0001] This invention belongs to the field of reliability analysis of high-end complex equipment, and specifically relates to a method and system for reliability analysis of high-end complex equipment. Background Technology

[0002] Reliability, as one of the six key general quality characteristics of high-end complex equipment, is closely related to its safe operation. Reliability analysis is a crucial aspect of identifying potential risks and failures in equipment and ensuring its safe operation.

[0003] FMEA (Failure Mode and Effect Analysis) is a mainstream reliability analysis method, initially used in the early 1950s in the United States for the design and analysis of a fighter jet operating system. In the mid-1960s, FMEA technology was officially used in the US aerospace industry. In 1976, the US Department of Defense released military standards for FMEA, but only for design purposes. In the late 1970s, FMEA technology began to enter the automotive and medical device industries. In the early 1980s, it entered the microelectronics industry. In the mid-1980s, the automotive industry began using FMEA to validate its manufacturing processes. By 1988, the US Federal Aviation Administration (FAA) issued an advisory requiring the use of FMEA in the design and analysis of all aerospace systems. In 1991, ISO 9000 recommended the use of FMEA to improve product and process design. In 1994, FMEA became a certification requirement for QS-9000 and one of the five core tools of QS-9000. To date, FMEA has developed into a scientific and complete analytical method in engineering practice and has become an essential method for continuous quality improvement in traditional industries and high-tech industries. At the same time, it has also been adopted by other methods as a common reliability analysis concept.

[0004] The advantages of traditional FMEA are its simplicity, ease of implementation, and ease of mastery and promotion. Even without data, anyone with relevant engineering experience can perform this task. Therefore, it is inexpensive, yields good results, and is widely welcomed by engineering and technical personnel.

[0005] However, as equipment functions and structures become increasingly complex, traditional FMEA has revealed some shortcomings in risk analysis, failing to fully meet the demands of increasingly complex equipment reliability analysis. Traditional FMEA does not consider the relative weights of the three risk factors: S (Safety), O (Occupation), and D (Defect). Different combinations of S, O, and D may result in the same risk priority number, but their underlying risk implications can be completely different, hindering decision-makers from taking appropriate countermeasures, easily leading to wasted resources and potentially failing to achieve the desired outcome. Furthermore, these three risk factors are often difficult to determine, making FMEA teams highly subjective in their scoring, lacking universal, accurate, and detailed scoring criteria. FMEA assessments suffer from significant uncertainty. Finally, traditional FMEA does not consider the correlation between different failure modes, as the severity and incidence of a particular failure mode may be influenced by other failure modes, necessitating the inclusion of failure propagation in the analysis. Summary of the Invention

[0006] The purpose of this invention is to provide a reliability analysis method and system for high-end complex equipment, so as to overcome the problems of insufficient analytical and expressive capabilities of traditional FMEA models and high uncertainty in expert evaluation in the current field of reliability analysis of high-end complex equipment.

[0007] A reliability analysis method for high-end complex equipment includes the following steps:

[0008] S1, establish the correlation strength data and risk factor weight vector between various complex equipment failures based on typical failure mode data;

[0009] S2, Construct a risk assessment matrix based on typical fault data of complex equipment and the correlation strength data between faults;

[0010] S3. Based on the weight vector of risk factors and the risk assessment matrix, calculate the risk benefit value of each failure mode, and finally calculate the risk priority ranking of each failure mode.

[0011] Preferably, the typical fault-related data report includes typical fault mode data of complex equipment, monitoring data returned by monitoring points established to detect these fault modes, and the average maintenance time for each fault mode.

[0012] Preferably, based on the failure mode data, an n×n failure mode autocorrelation matrix is ​​constructed, and the constructed correlation strength data matrix is ​​shown below:

[0013]

[0014] Among them, R m Let a be the domino matrix given by the m-th expert. ita represents the correlation between the i-th failure mode and the t-th failure mode. ti This represents the correlation between the t-th fault mode and the i-th fault mode.

[0015] Preferably, a weighted vector is established for risk factors: correlation, hazard, and failure warning rate, as shown below:

[0016]

[0017] W m It is the risk factor assessment vector given by the m-th expert. Let m be the weight given by the m-th expert to the correlation among risk factors. Let m be the weight given by the m-th expert to the harm in the risk factors. This represents the weight given by the m-th expert to the failure warning rate among risk factors.

[0018] Preferably, based on the correlation strength data matrix between multiple failure modes, the domino matrix is ​​integrated according to the following formula:

[0019] The correlation risk factor for each failure mode is measured using the following formula:

[0020]

[0021] in A vector of size 1×n with all elements equal to 1. The final correlation vector is obtained, where the i-th element represents the measurement result of the correlation of the i-th failure mode as a risk factor.

[0022] Preferably, the severity of a failure is measured by the economic loss caused by each failure mode. The economic loss caused by a failure is linearly positively correlated with the average downtime for maintenance. The severity of a failure is measured by the average maintenance time.

[0023] Preferably, the warning rate metric for each fault mode is calculated:

[0024] The improved Taguchi process capability index (Cpm-R) was used to statistically analyze the monitoring data returned from the monitoring points corresponding to each failure mode. The calculation was performed once for every 20 points. The improved Taguchi process capability index is shown in the following formula:

[0025]

[0026] Let z k(k = 1, 2, ..., n) is a set of data, n is the number of data, USL is the upper limit of the specification, LSL is the lower limit of the specification, μ is the mean, σ is the standard deviation, M is the specification center, T is the control target, and d is the width of half the specification.

[0027] Preferably, based on a threshold X set within the evaluation system, the calculated improved Taguchi process capability index is divided into two parts: the part exceeding the threshold and the part within the threshold range; the fault warning rate is calculated according to the following formula:

[0028]

[0029] Where EW is the warning rate for a certain fault mode, P is the number of Cpm-Rs exceeding the threshold X, and N is the total number of Cpm-Rs.

[0030] Preferably, the following formula is used for standardization to facilitate the next step of calculation:

[0031] The risk priority of each failure mode is calculated using a multi-objective optimization method based on full multiplication proportional analysis.

[0032] A reliability analysis system for high-end complex equipment, comprising a preprocessing module and an analysis module;

[0033] The preprocessing module is used to establish the correlation strength data between various complex equipment failures and the weight vector of risk factors based on typical failure mode data; and to construct a risk assessment matrix based on the relevant data of typical failures of complex equipment and the correlation strength data between failures.

[0034] The analysis module calculates the risk-benefit value of each failure mode based on the weight vector of risk factors and the risk assessment matrix, and finally calculates the risk priority ranking of each failure mode.

[0035] Compared with the prior art, the present invention has the following beneficial technical effects:

[0036] This invention provides a reliability analysis method for high-end complex equipment. It establishes correlation strength data and risk factor weight vectors among various complex equipment failures based on typical failure mode data. A risk assessment matrix is ​​constructed based on typical failure correlation data and correlation strength data between failures. The risk benefit value of each failure mode is calculated based on the risk factor weight vector and the risk assessment matrix. This method reduces human error and uncertainty in the reliability analysis process, effectively improving the objectivity and accuracy of reliability analysis for complex equipment. Based on a failure autocorrelation matrix, this invention incorporates the correlation between different failures and proposes a method for measuring the correlation degree, solving the problem that traditional FMEA does not consider the correlation between different failure modes.

[0037] This invention introduces the concept of multi-criteria decision-making, incorporating the weights of different risk factors to better prioritize failure modes. This invention can provide product optimization suggestions for reliability design departments, reliability control point configuration suggestions for manufacturing enterprises, and some of its processes offer new technical approaches for reliability analysis of other similar products. It can also be applied to risk management decisions in other areas of quality and reliability engineering.

[0038] The process capability index is used to statistically analyze a large amount of monitoring data generated by sensors that detect failure modes. Three risk factors, namely correlation, severity, and failure warning rate, are established to address the problem that traditional FMEA does not take into account the correlation between failure modes, which can meet the reliability analysis needs of high-end complex equipment. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of a reliability analysis method for high-end complex equipment in an embodiment of the present invention. Detailed Implementation

[0040] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0041] like Figure 1 As shown, a reliability analysis method for high-end complex equipment includes the following steps:

[0042] S1, Fault Data Collection Steps: Collect the average maintenance time of typical fault modes based on historical records; set up monitoring points for monitoring typical fault modes and return monitoring data; establish the correlation strength data between faults of various complex equipment and the weight vector of risk factors;

[0043] 1.1) Collection of typical fault-related data for complex equipment: Collect and organize data on typical fault modes of n complex equipment, as well as monitoring data returned by monitoring points set up to detect these fault modes, and the average maintenance time for each fault mode.

[0044] 1.2) Determine the correlation strength data between various failure modes; based on the failure mode data collected in 1.1), obtain an n×n failure mode autocorrelation matrix. This matrix determines the dependencies between different failure modes and expresses the relationships between different failure modes in a bidirectional manner. The constructed correlation strength data matrix is ​​shown below:

[0045]

[0046] Among them, R m Let a be the domino matrix given by the m-th expert. it a represents the correlation between the i-th failure mode and the t-th failure mode. ti a represents the correlation between the t-th failure mode and the i-th failure mode. it and a ti The values ​​are obtained from Table 1 below:

[0047] Table 1 shows the correlation between different failure modes.

[0048]

[0049]

[0050] Calculate the weight vector for each risk factor:

[0051] A weight vector is established for the three risk factors designed in this invention: correlation (R), hazard (H), and early warning rate (EW), as shown below:

[0052]

[0053] Among them, W m It is the risk factor assessment vector given by the m-th expert. Let m be the weight given by the m-th expert to the correlation among risk factors. Let m be the weight given by the m-th expert to the harm in the risk factors. This represents the weight given by the m-th expert to the failure warning rate among risk factors.

[0054] The risk factor weight vectors from the expert assessments are integrated according to the following formula:

[0055]

[0056] In the formula, W m Let W be the risk factor weight vector given by the m-th expert, k be the number of evaluation experts, and W be [ω]. R ω H ω EW This is a preliminary integration of the risk factor weight vectors obtained from all expert assessments.

[0057] Finally, the normalized weight vector of the risk factors is obtained using the following method:

[0058]

[0059] The final risk factor weight vector is

[0060] S2, Fault Assessment Matrix Construction Steps: Based on the typical fault-related data of complex equipment and the correlation strength data between faults in step S1, construct a risk assessment matrix;

[0061] 2.1) Calculate the correlation strength data for each failure mode:

[0062] Based on the correlation strength data matrix among multiple failure modes constructed in 1.2), the domino matrix is ​​integrated according to the following formula:

[0063]

[0064] The correlation risk factor for each failure mode is measured using the following formula:

[0065]

[0066] in A vector of size 1×n with all elements equal to 1. The final correlation vector is obtained, and the i-th element in the vector is the measurement result of the risk factor correlation of the i-th failure mode.

[0067] 2.2) Calculate the hazard measurement results for each failure mode:

[0068] This invention is based on cost-benefit analysis, measuring the severity of a failure by the economic losses incurred after each failure mode occurs, with mean time to repair (MTR) representing average downtime for maintenance. Since there is a linear positive correlation between the economic losses caused by a failure and the MTR, this invention uses MTR to measure the severity of a failure.

[0069] 2.3) Calculate the warning rate metric for each fault mode:

[0070] The improved Taguchi process capability index (Cpm-R) was used to statistically analyze the monitoring data returned from the monitoring points corresponding to each failure mode, with calculations performed every 20 points. The improved Taguchi process capability index is shown in the following formula.

[0071]

[0072] Let z k (k = 1, 2, ..., n) is a set of data, where n is the number of data points. USL is the upper limit of the specification, LSL is the lower limit of the specification, μ is the mean, σ is the standard deviation, M is the specification center, T is the control target, and d is the width of half the specification.

[0073]

[0074] The corresponding evaluation criteria are shown in Table 2 below.

[0075] Table 2

[0076]

[0077]

[0078] Then, based on the threshold X set within the evaluation system, it is considered that when the equipment's operating condition exceeds X, it reaches a level where a warning needs to be issued to the regulator. The calculated improved Taguchi process capability index is divided into two parts using the threshold: the part exceeding the threshold and the part within the threshold range. Finally, the failure warning rate is calculated using the following formula:

[0079]

[0080] Where EW is the warning rate for a certain fault mode, P is the number of Cpm-Rs exceeding the threshold X, and N is the total number of Cpm-Rs.

[0081] S3, Failure Mode Risk Priority Ranking Steps: Based on the risk factor weight vector in S1 and the risk assessment matrix established in S2, calculate the risk benefit value of each failure mode, and finally calculate the risk priority ranking of each failure mode.

[0082] 3.1) Matrix standardization:

[0083] Since the three risk factors come from different dimensions when constructing the risk assessment matrix, they are standardized according to the following formula to facilitate the next step of calculation:

[0084]

[0085] 3.2) The risk priority of each failure mode is calculated using a multi-objective optimization method based on full multiplication proportional analysis:

[0086] 3.2.1) Constructing a ratio system

[0087] The risk-benefit value of each failure mode under the ratio system is calculated using the following formula.

[0088]

[0089] Where g represents the number of benefit-type indicators, and (ng) represents the number of cost-type indicators. The optimal solution based on the ratio system has the greatest benefit y. i The ranking of these methods is in descending order:

[0090]

[0091] 3.2.2) Constructing a reference level model

[0092] First, find the worst or best performing option among each risk factor indicator, as shown in the following formula.

[0093]

[0094] Benefit values ​​z of each scheme calculated based on the reference level model i The calculation method is shown in the following formula.

[0095]

[0096] The optimal solution based on the reference level model method has the minimum utility value z. i The methods are ranked in ascending order:

[0097]

[0098] 3.2.3) Constructing the complete multiplication form

[0099] The risk-benefit values ​​for each failure mode in the fully multiplicative form are calculated as shown in the following formula.

[0100]

[0101] The optimal solution based on the full multiplicative form has the maximum utility value u. i This method ranks items in descending order:

[0102]

[0103] 3.2.4) Ranking Aggregation

[0104] Taking into account the position of each option in each sub-ranking method, the final failure mode risk priority ranking is calculated as shown in the following formula.

[0105] RPM(A i )=1 / (1 / r(y i )+1 / r(z i )+1 / r(u i ))

[0106] Where r(y) i ), r(z i ) and r(u i These are the ratio system, the reference level model, and the ranking in fully multiplicative form, respectively. The optimal scheme based on this ranking aggregation method has the minimum RPM (A). i )value.

[0107] The final results represent the risk priority of each failure mode, realizing the risk analysis function in traditional FMEA, and also realizing the reliability analysis of equipment.

[0108] This invention significantly addresses the significant uncertainty inherent in expert assessments during traditional FMEA (Factor-Driven Engineering Analysis), particularly regarding hazard severity and warning rates. It proposes objective measurement methods based on cost-benefit analysis and an improved Taguchi process capability index, reducing human error and uncertainty in reliability analysis, and effectively improving the objectivity and accuracy of reliability analysis for complex equipment. The model, based on a fault autocorrelation matrix, incorporates the correlations between different faults and proposes a method for measuring the degree of correlation, solving the problem of traditional FMEA failing to consider the correlations between different failure modes. This invention introduces the concept of multi-criteria decision-making, utilizing the MULTIMOORA method to incorporate the weights of different risk factors, better prioritizing failure modes. This invention can provide product optimization suggestions for reliability design departments, reliability control point configuration suggestions for manufacturing enterprises, and some processing steps offer new technical approaches for reliability analysis of other similar products. It can also be applied to risk management decisions in other areas of quality and reliability engineering.

Claims

1. A reliability analysis method for high-end complex equipment, characterized in that, Includes the following steps: S1, establish the correlation strength data and risk factor weight vector between various complex equipment failures based on typical failure mode data; S2, Construct a risk assessment matrix based on typical fault data of complex equipment and the correlation strength data between faults; S3. Based on the weight vector of risk factors and the risk assessment matrix, calculate the risk benefit value of each failure mode, and finally calculate the risk priority ranking of each failure mode. Calculate the warning rate metric for each failure mode: Using the improved Taguchi process capability index Statistical analysis was performed on the monitoring data returned from the monitoring points corresponding to each failure mode, with calculations performed every 20 points. The improved Taguchi process capability index is shown in the following formula: It is the mean. T is the standard deviation, and T is the control target. It is half the width of the specification. Based on the threshold X set within the evaluation system, the calculated improved Taguchi process capability index is divided into two parts: the part exceeding the threshold and the part within the threshold range. The fault warning rate is calculated according to the following formula: Where EW is the warning rate for a certain fault mode, and P is the rate at which the threshold X is exceeded. The number of items, N is the total number of items. The number of; Standardize risk factors: The risk priority of each failure mode is calculated using a multi-objective optimization method based on full multiplication proportional analysis. 3.2.1) Calculate the risk-benefit value of each failure mode under the ratio system according to the following formula: Where g represents the number of benefit-related indicators and ng represents the number of cost-related indicators; the optimal solution based on the ratio system has the greatest benefit. The ranking of these methods is in descending order: 3.2.2) Obtain the worst-performing or best-performing solution for each risk factor indicator: Benefit values ​​of each scheme calculated based on the reference level model The calculation method is shown in the following formula; The optimal solution based on the reference level model method has the minimum utility value. The methods are ranked in ascending order: 3.2.3) The risk-benefit values ​​for each failure mode in the fully multiplicative form are calculated as follows: The optimal solution based on the complete multiplicative form has the maximum utility value. This method ranks items in descending order: 3.2.4) Taking into account the position of each option in each sub-ranking method, the final failure mode risk priority ranking is calculated as follows: in , and These are the ratio system, the reference level model, and the ranking in a fully multiplicative form, respectively. The optimal solution based on this ranking aggregation method has the minimum... value.

2. The reliability analysis method for high-end complex equipment according to claim 1, characterized in that, Typical failure-related data include data on typical failure modes of complex equipment, as well as monitoring data returned by monitoring points established to detect these failure modes, and the average maintenance time for each failure mode.

3. The reliability analysis method for high-end complex equipment according to claim 1, characterized in that, Based on failure mode data, establish The failure mode autocorrelation matrix is ​​constructed as follows: in, Let m be the domino matrix given by the m-th expert. This represents the correlation between the i-th fault mode and the t-th fault mode. This represents the correlation between the t-th fault mode and the i-th fault mode.

4. The reliability analysis method for high-end complex equipment according to claim 1, characterized in that, Establish a weighted vector for risk factors: correlation, hazard, and failure warning rate, as shown below: It is the risk factor assessment vector given by the m-th expert. Let m be the weight given by the m-th expert to the correlation among risk factors. Let m be the weight given by the m-th expert to the harm in the risk factors. This represents the weight given by the m-th expert to the failure warning rate among risk factors.

5. The reliability analysis method for high-end complex equipment according to claim 3, characterized in that, Based on the correlation strength data matrix among multiple failure modes, the domino matrix is ​​integrated according to the following formula: k represents the number of assessment experts. The correlation risk factor for each failure mode is measured using the following formula: in for A vector in which all elements are 1. The final correlation vector is obtained, and the i-th element in the vector is the measurement result of the risk factor correlation of the i-th failure mode.

6. The reliability analysis method for high-end complex equipment according to claim 1, characterized in that, The severity of a failure is measured by the economic losses caused by each failure mode. The economic losses caused by a failure are linearly positively correlated with the average downtime for maintenance. The severity of a failure is measured by the average maintenance time.

7. A high-end complex equipment reliability analysis system for the method of claim 1, characterized in that, Preprocessing module and analysis module; The preprocessing module is used to establish the correlation strength data and risk factor weight vectors between various complex equipment failures based on typical failure mode data. A risk assessment matrix is ​​constructed based on typical failure data of complex equipment and the correlation strength data between failures; The analysis module calculates the risk-benefit value of each failure mode based on the weight vector of risk factors and the risk assessment matrix, and finally calculates the risk priority ranking of each failure mode.

Images