Equipment health assessment method, system, and equipment based on multidimensional data fusion

Abstract

This invention belongs to the field of equipment health management technology, specifically involving a method, system, and equipment for evaluating equipment health based on multi-dimensional data fusion. The method constructs a multi-level evaluation system, uses a degradation degree function to normalize multi-source data, and proposes a dynamic weight correction mechanism based on fluctuation characteristics. Specifically, it selects the appropriate degradation degree function to calculate the degradation degree value based on the indicator characteristics; determines the initial combined weights using the correlation coefficient method and the analytic hierarchy process; corrects the weights by constructing a risk enhancement factor using the coefficient of variation; dynamically adjusts the weights of the first-level indicators based on the weighted degradation degree and average volatility; calculates the membership degree of each health level using different distribution functions; and finally determines the equipment health level through a multi-level evaluation matrix. This invention solves the problems of inaccurate static weight allocation and insufficient multi-source data fusion in existing technologies, achieving accurate and dynamic assessment of equipment health status.

Description

Technical Field

[0001] This invention belongs to the field of equipment health management technology, specifically relating to a method, system, and equipment for evaluating equipment health based on multi-dimensional data fusion. Background Technology

[0002] Currently, equipment health assessment mainly employs threshold-based judgment methods or static weighted comprehensive evaluation models based on a single data source. Existing technical solutions typically achieve this through two methods: first, a fixed threshold alarm mechanism based on equipment operating parameters, such as triggering an alarm when the temperature exceeds a preset threshold; second, constructing an evaluation system using static weight allocation methods such as the Analytic Hierarchy Process (AHP) and combining it with expert scoring for health assessment. For example, the health assessment method for coal-fired power plant equipment disclosed in patent CN115619107A uses weighted calculations based on fixed-dimensional evaluation indicators such as safety and stability. Furthermore, some improved solutions attempt to introduce fuzzy theory to handle evaluation uncertainties; for instance, patent CN112257984A uses a fuzzy comprehensive evaluation method to classify the status of power equipment. These methods can reflect the equipment operating status to some extent, but they still have significant limitations.

[0003] Existing equipment health assessment methods suffer from the following technical shortcomings: First, static weight allocation methods cannot adapt to the dynamic changes in equipment operating status, leading to discrepancies between assessment results and actual health conditions. Second, traditional methods lack the ability to integrate multi-source heterogeneous data, making it difficult to effectively uncover the correlations between data from multiple dimensions such as vibration, temperature, and environment. Third, existing solutions lack dynamic response mechanisms to the fluctuation characteristics of indicators, failing to identify potential fault risks in a timely manner. For example, when equipment experiences intermittent anomalies, fixed threshold methods are prone to false alarms or missed alarms, while static weight models struggle to adjust the weight ratios of key indicators in a timely manner. Furthermore, existing health level classification standards rely too heavily on subjective experience and lack objective, quantitative evaluation criteria, resulting in a lack of comparability between assessment results from different devices. These problems severely restrict the accuracy and practicality of equipment health management systems. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, and device for evaluating equipment health based on multi-dimensional data fusion, so as to solve the technical problems of inaccurate evaluation results and poor dynamic adaptability caused by insufficient static weight allocation and multi-source data fusion in existing equipment health evaluation methods.

[0005] The present invention achieves the above objectives through the following technical solutions: Firstly, this invention proposes a method for evaluating device health based on multi-dimensional data fusion, the method comprising the following steps: S1. Construct an evaluation system that includes multiple primary indicator subsystems, each of which contains several secondary indicators. S2. Normalize the pre-collected secondary indicator data and calculate the degradation value of each secondary indicator. S3. Calculate the initial combination weights of each secondary indicator, and correct the initial combination weights based on the fluctuation characteristics of each secondary indicator to obtain the final weights of the secondary indicators; dynamically calculate the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem. S4. Calculate the membership degree of each secondary indicator to each preset health level based on the membership function, and construct the membership degree matrix; S5. Based on the final weights and membership matrices of the secondary indicators, calculate the evaluation vector for each primary indicator; combine the dynamically adjusted weights of the primary indicators to calculate the final evaluation vector. S6. Based on the final evaluation vector, determine the health level of the device according to the principle of maximum membership.

[0006] Furthermore, before calculating the degradation value of each secondary indicator, the method further includes selecting a fitness function according to the data type of the secondary indicator, specifically including: For the first target indicator, the first degradation function is selected as follows: ; For the second target index, the second degradation function is selected as follows: ; For the third target indicator, the third degradation function is selected as follows: ; Among them, the smaller the value of the first target indicator, the better the equipment status; the value of the second target indicator is within the preset target range, indicating that the equipment status is optimal; and the larger the value of the third target indicator, the better the equipment status. The measured value of the indicator. The minimum value of the indicator. The maximum value of the indicator. This represents the lower limit of the target parameter range. This indicates the upper limit of the target parameter range.

[0007] Furthermore, in step S3, the calculation of the final weight of the secondary indicator includes: S301. The correlation coefficient method is used to calculate the correlation coefficient between the secondary indicators of each operating status and the health outcome indicators, and to obtain the objective weights. The correlation coefficient method uses the Pearson correlation coefficient, and the health outcome index is the expert score or the normalized value of the set performance parameter in the equipment's historical operation. S302. Construct a judgment matrix using the analytic hierarchy process (AHP) and calculate the subjective weights of each secondary indicator. ; S303, Comprehensive objective weights through weighted average and subjective weight To obtain the initial combined weights As shown in the following formula: ;in, This represents the objective weight of the j-th indicator. This represents the subjective weight of the j-th indicator. It is an adjustment parameter, with a value range of [0,1]. S304. Calculate the coefficient of variation of each secondary indicator within the preset time window. As shown in the following formula: ,in and This represents the standard deviation and mean of the secondary indicators within a preset time window. The volatility of a quantitative indicator is indicated by its value; the larger the value, the stronger the volatility of that secondary indicator. S305, Based on the aforementioned coefficient of variation Constructing risk enhancement factors As shown in the following formula: ;in, To amplify the upper limit of risk, The value range is set to ; S306. Utilizing the aforementioned risk enhancement factor For the initial combined weights Make corrections to obtain the corrected combined weights. As shown in the following formula: ; S307. Adjusted combined weights Normalization is performed to obtain the final weights of the secondary indicators. .

[0008] Furthermore, the dynamic calculation of the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem includes: S311. Calculate the weighted degradation degree of each first-level indicator subsystem. As shown in the following formula: ,in, Indicates the first Normalized degradation values ​​of each secondary indicator; S312. Calculate the average volatility of each first-level indicator subsystem. As shown in the following formula: ,in, Indicates the first Average volatility of each subsystem Representation Subsystem The number of secondary indicators included. Indicators coefficient of variation; S313, Based on weighted degradation and average volatility Calculate the comprehensive risk factor As shown in the following formula: ,in, Indicates the first Comprehensive risk factors for each subsystem represents the risk adjustment factor, indicating the weight preference for the degree of deterioration, with a value range of [0, 1]. S314. Based on comprehensive risk factors Dynamically allocate the weights of primary indicators As shown in the following formula: .

[0009] Furthermore, the step of calculating the membership degree of each secondary indicator to each preset health level based on the membership function includes: substituting the deterioration value of each secondary indicator into the following four membership functions: A descending ridge distribution function is selected to calculate its membership degree to the first health level; an intermediate ridge distribution function is selected to calculate its membership degree to the second and third health levels; and an ascending ridge distribution function is selected to calculate its membership degree to the fourth health level. The health levels are defined in order of the degree of equipment condition: the first health level corresponds to the equipment being in the best operating condition; the second health level corresponds to the equipment being in the normal operating condition; the third health level corresponds to the equipment being in a state requiring attention; and the fourth health level corresponds to the equipment being in a faulty or imminent faulty state.

[0010] Furthermore, the evaluation vector for each primary indicator is calculated. As shown in the following formula: ; in, This is the final weight vector for the secondary indicators. Let represent the membership matrix of the i-th first-level index subsystem.

[0011] Furthermore, the final evaluation vector B is calculated by combining the dynamically adjusted weights of the primary indicators, as shown in the following formula: ; In the formula This is the weight vector for the primary indicators. The evaluation matrix is ​​composed of the evaluation vectors of the primary indicators.

[0012] Secondly, this invention proposes a device health assessment system based on multi-dimensional data fusion, used to implement the device health assessment method described above. The system includes: The data acquisition module is used to construct an evaluation system that includes multiple primary indicator subsystems, each of which contains several secondary indicators. The first calculation module is used to normalize the pre-collected secondary indicator data and calculate the degradation value of each secondary indicator. The second calculation module is used to calculate the initial combination weights of each secondary indicator, correct the initial combination weights based on the fluctuation characteristics of each secondary indicator, and obtain the final weights of the secondary indicators; and dynamically calculate the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem. The third calculation module is used to calculate the membership degree of each secondary indicator to each preset health level based on the membership function, and to construct the membership degree matrix; The fourth calculation module is used to calculate the evaluation vector of each primary indicator based on the final weights and membership matrix of the secondary indicators; and to calculate the final evaluation vector by combining the dynamically adjusted weights of the primary indicators. The health assessment module is used to determine the health level of the device based on the principle of maximum membership, by combining the final evaluation vector.

[0013] Thirdly, the present invention proposes an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the device health evaluation method based on multidimensional data fusion as described above.

[0014] Fourthly, the present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the device health evaluation method based on multidimensional data fusion as described above.

[0015] The beneficial effects of this invention are as follows: This invention effectively improves the accuracy and adaptability of equipment condition monitoring by constructing a multi-dimensional, dynamically integrated health evaluation system. Its core lies in establishing a multi-level evaluation architecture encompassing operating status, historical data, environmental parameters, and equipment attributes. It determines indicator weights through a combination of correlation coefficient and analytic hierarchy process (AHP) methods and introduces a dynamic correction mechanism based on fluctuation characteristics, enabling the evaluation model to automatically adapt to changes in equipment condition. This technical solution uses a degradation degree function to standardize multi-source data and a membership function to achieve objective quantitative assessment of health levels. By dynamically adjusting the weights of primary and secondary indicators, the system significantly improves its sensitivity to abnormal equipment conditions, enabling earlier identification of potential fault risks. Compared to existing technologies, this invention overcomes the limitations of static weight allocation and solves the problem of insufficient multi-source data fusion, making the evaluation results more closely reflect the actual operating conditions of the equipment. Attached Figure Description

[0016] Figure 1 A flowchart illustrating a device health evaluation method based on multidimensional data fusion provided in a specific embodiment of this application; Figure 2 This is another flowchart illustrating a device health evaluation method based on multidimensional data fusion provided in a specific embodiment of this application; Figure 3 This is another flowchart illustrating a device health evaluation method based on multidimensional data fusion provided in a specific embodiment of this application; Figure 4 This is a schematic diagram of the membership function of a device health evaluation method based on multidimensional data fusion provided in a specific embodiment of this application. Detailed Implementation

[0017] The present application will now be described in further detail with reference to the accompanying drawings. It should be noted that the following specific embodiments are only used to further illustrate the present application and should not be construed as limiting the scope of protection of the present application. Those skilled in the art can make some non-essential improvements and adjustments to the present application based on the above application content.

[0018] Example 1 Please combine Figure 1 , Figure 2 and Figure 3 In one specific embodiment, this application proposes a device health evaluation method based on multi-dimensional data fusion, the method comprising the following steps: S1. Construct an evaluation system that includes multiple primary indicator subsystems, each of which contains several secondary indicators. S2. Normalize the pre-collected secondary indicator data (corresponding to the specific data of the above secondary indicators) and calculate the degradation value of each secondary indicator. S3. Calculate the initial combination weights of each secondary indicator, and correct the initial combination weights based on the fluctuation characteristics of each secondary indicator to obtain the final weights of the secondary indicators; dynamically calculate the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem. S4. Calculate the membership degree of each secondary indicator to each preset health level based on the membership function, and construct the membership degree matrix; S5. Based on the final weights and membership matrices of the secondary indicators, calculate the evaluation vector for each primary indicator; combine the dynamically adjusted weights of the primary indicators to calculate the final evaluation vector. S6. Combine the final evaluation vector and determine the health level of the equipment according to the principle of maximum membership.

[0019] It should be noted that the health level in this application refers to different health states from "excellent" to "faulty", specifically including "excellent", "good", "average" and "faulty".

[0020] It is understood that the purpose of this application is to provide a dynamic evaluation method for equipment health based on multi-dimensional data fusion. To achieve this method, embodiments of the present invention construct a multi-level evaluation model that includes operating status, historical data, environmental parameters, and equipment attributes. A dynamic weight adjustment mechanism and a fuzzy comprehensive evaluation algorithm are used to achieve accurate assessment of equipment health status. In practical applications, equipment health assessment is usually conducted by maintenance personnel based on experience or fixed threshold rules. However, traditional methods are difficult to cope with dynamic changes under complex operating conditions and cannot effectively integrate multi-source heterogeneous data. Therefore, embodiments of the present invention combine data fusion technology with a dynamic weight adjustment mechanism and introduce a risk enhancement factor based on fluctuation characteristics, thereby achieving intelligent and accurate assessment of equipment health status.

[0021] In practical implementation, for the evaluation system comprising multiple primary indicator subsystems constructed in step S1, the composition of the primary indicator subsystems needs to be determined first based on the equipment type and monitoring requirements. Taking industrial rotating equipment as an example, this embodiment divides the evaluation system into four primary indicator subsystems: the operating status subsystem includes real-time monitoring indicators such as vibration, temperature, current, and voltage; the historical data subsystem includes maintenance record indicators such as maintenance frequency, failure rate, and mean time between failures (MTBF); the environmental parameter subsystem covers external environmental indicators such as ambient temperature, humidity, and dust concentration; and the equipment attribute subsystem includes inherent characteristic indicators such as design life and material durability. Each primary subsystem has several secondary indicators. For example, the vibration indicator in the operating status subsystem can be further subdivided into three sub-indicators: horizontal vibration, vertical vibration, and axial vibration.

[0022] It is important to note that the selection of indicators for each primary subsystem must meet the following principles: (1) comprehensiveness, covering the main influencing factors of equipment health status; (2) measurability, indicator data can be obtained through sensors or system recording; (3) independence, the correlation between indicators should be minimized. During implementation, an indicator database needs to be established to store basic information such as parameter type (smaller is better, intermediate, or larger is better), unit of measurement, and safety threshold for each indicator, providing data support for subsequent degradation calculations and weight allocation.

[0023] In step S2, the pre-collected secondary indicator data are normalized, and the degradation degree value of each secondary indicator is calculated. The relative degradation degree analysis method is used to normalize these indicators. The relative degradation degree refers to the degree of degradation in the current actual state compared to the degree of degradation under fault conditions. The relative degradation degree ranges from [0,1]. There are three main methods for calculating the degradation degree of indicator parameters. For example, for indicator parameters where a smaller value is better, such as the equipment temperature, the overall parameter of its degradation degree function needs to be estimated to have a maximum value. and minimum value .

[0024] More specifically, as a preferred approach, before calculating the degradation values ​​of each secondary indicator, the method further includes selecting a fitness function based on the data type of the secondary indicator. Specifically, for the first target indicator, i.e., for indicators where smaller values ​​are better, such as equipment temperature, the overall parameter of its degradation function needs to be estimated to have a maximum value. and minimum value The first degradation function is selected as follows: ; In the formula The measured value of the indicator. This is the minimum value of the indicator (usually the minimum value in historical data or a set safety lower limit). Given the maximum value of the indicator (usually the maximum value in historical data or a set alarm limit), this function indicates that when the indicator value is below... The degradation degree is 0 when the value is in and Between these values, the degradation increases linearly from 0 to 1; when the value is higher than 1... At that time, the degradation level is 1.

[0025] For the second target indicator, namely intermediate indicator parameters such as equipment voltage, the overall parameter of its degradation function needs to be estimated to have a maximum value. Minimum value and the optimal range of the variable [ , The second degradation function is selected as follows: ; In the formula The measured value of the indicator. This is the minimum value of the indicator (usually the minimum value in historical data or a set safety lower limit). This represents the lower limit of the optimal range. This represents the upper limit of the optimal range. This is the maximum value of the indicator (usually the maximum value in historical data or a set alarm limit). This function indicates when the indicator value is within the optimal range... , The degradation level is 0 within the specified time; when the value is lower than 0. But not less than When the value is higher than 1, the degradation degree increases linearly from 0 to 1; when the value is higher than 1, the degradation degree increases linearly from 0 to 1. But not higher than When the value is below 1, the degradation increases linearly from 0 to 1; when the value is below 1... or higher At that time, the degradation level is 1.

[0026] For the third objective metric, namely, a larger value is always better, such as the mean time between failures (MBTF) of an equipment, the overall parameter of its degradation function needs to be estimated to have a maximum value. and minimum value The third degradation function is selected as follows: ; In the formula The measured value of the indicator. The maximum value of the indicator. This function represents the minimum value of the indicator, meaning that when the indicator value is not lower than... The degradation degree is 0 when the value is in and Between, the degree of degradation increases The value decreases and increases linearly from 0 to 1; when the value is not higher than... At that time, the degradation level is 1.

[0027] Among them, the smaller the value of the first target indicator, the better the equipment status; the value of the second target indicator being within the preset target range indicates the optimal equipment status; and the larger the value of the third target indicator, the better the equipment status.

[0028] As a preferred option, in step S3, the calculation of the final weights of the secondary indicators includes: S301. The correlation coefficient method is used to calculate the correlation coefficient between the secondary indicators of each operating status and the health outcome indicators, and to obtain the objective weights. The correlation coefficient method uses the Pearson correlation coefficient, and the health outcome indicators are the expert scores or normalized values ​​of the set performance parameters in the equipment's historical operation.

[0029] The health outcome indicators are the health scores among the various parameters of the equipment's operating status, determined based on historical data and expert evaluation.

[0030] S302. Construct a judgment matrix using the analytic hierarchy process (AHP) and calculate the subjective weights of each secondary indicator. ; S303, Comprehensive objective weights through weighted average and subjective weight To obtain the initial combined weights As shown in the following formula: ;in, This represents the objective weight of the j-th indicator (calculated using the correlation coefficient method). This represents the subjective weight of the j-th indicator (calculated using the analytic hierarchy process). It is an adjustment parameter, with a value range of [0,1]. S304. Calculate the coefficient of variation of each secondary indicator within the preset time window. As shown in the following formula: ,in and This represents the standard deviation and mean of the secondary indicators within a preset time window. The volatility of a quantitative indicator is indicated by its value; the larger the value, the stronger the volatility of that secondary indicator. S305, Based on Coefficient of Variation Constructing risk enhancement factors As shown in the following formula: ;in, The risk amplification upper limit is an artificially set maximum permissible multiple for the risk amplification effect of an indicator. It is used to limit the degree of risk enhancement caused by the volatility of the indicator. It is usually set directly by industry experts or operation and maintenance teams, taking into account equipment characteristics, failure consequences, and management experience (for example, wind power equipment is usually set at...). ). The value range is set to Risk enhancement factor The design aims to enhance the evaluation impact of high volatility indicators; S306, Utilizing Risk Enhancement Factors For the initial combined weights Make corrections to obtain the corrected combined weights. As shown in the following formula: ; S307. Adjusted combined weights Normalization is performed to obtain the final weights of the secondary indicators. .

[0031] More specifically, in step S301, calculating the correlation coefficient includes: calculating the correlation coefficient between each operational status indicator j and the health outcome indicator Y. The Pearson correlation coefficient is typically used to measure the degree of linear correlation between two continuous variables, as shown in the following formula: ; In the formula This represents the Pearson correlation coefficient between the j-th operational status indicator and the health outcome indicator. This represents the value of the j-th indicator in the i-th sample. This represents the average value of the j-th indicator across all n samples. This represents the value of the health outcome indicator in the i-th sample. This represents the average of the health outcome index across all n samples, where n represents the number of samples.

[0032] The sample can be understood as an indicator of the operating status over a period of time, specifically the values ​​of vibration, temperature, voltage, and current.

[0033] Calculating the strength of correlation includes: correlation coefficient The range of values ​​for is [−1, +1]. The positive and negative signs indicate the direction of correlation, and the absolute value... Indicates the correlation strength. The closer the value is to 1, the stronger the correlation between the indicator and health outcomes.

[0034] In an alternative embodiment, this application also includes calculating an objective weight based on the correlation coefficient, as follows: ; In the formula This represents the objective weight of the j-th operational status indicator. This represents the absolute value of the correlation coefficient between the j-th indicator and the health outcome indicator, i.e., the strength of the correlation. This represents the sum of the absolute values ​​of the correlation coefficients of all m operational status indicators. The formula allocates weights proportionally, so that indicators with a stronger correlation to health outcomes receive higher objective weights.

[0035] More specifically, in step S302, the judgment matrix is ​​constructed as follows: A judgment matrix is ​​constructed based on expert experience to quantify the relative importance of each indicator, as shown in the following formula: ; In the formula This represents the judgment matrix, indicating the relative importance of each indicator, specifically the importance ratio of the i-th indicator to the j-th indicator. Values ​​are typically taken from 1 to 9 (Saaty proportional scaling method).

[0036] Calculating subjective weights: The subjective weights of each indicator are obtained by summing and normalizing the columns of the judgment matrix, as shown in the following formula: ; In the formula This represents the subjective weight of the j-th indicator. The numerator represents the value in the i-th row and j-th column of the judgment matrix, the denominator represents the sum of all elements in the j-th column, and the denominator represents the sum of all elements in the entire judgment matrix.

[0037] In an alternative embodiment, this application also includes a consistency check: since expert judgments may have some degree of inconsistency, a consistency check is needed to determine the credibility of the judgment matrix, as shown in the following formula: ; In the formula, the CR consistency ratio is used to evaluate the consistency of the judgment matrix. The largest eigenvalue of the judgment matrix is ​​the largest eigenvalue, which needs to be obtained through matrix calculation. The CI (Conformity Index) reflects the degree to which the judgment matrix deviates from consistency. The RI (Random Consistency Index) is the random consistency index, which can be obtained by looking up a table. This represents the order of the judgment matrix, i.e., the number of indicators. If CR < 0.10, the consistency of the judgment matrix is ​​considered acceptable. If CR ≥ 0.10, the judgment matrix is ​​considered inconsistent, and the expert's judgment needs to be adjusted.

[0038] Based on the above embodiments, this invention proposes a dynamic adaptive weighting mechanism that dynamically adjusts the weights of equipment health evaluation indicators by integrating objective data analysis and subjective expert experience. The technical solution combines the correlation coefficient method with the analytic hierarchy process (AHP), considering both the objective correlation between indicators and health status and incorporating expert domain knowledge. Furthermore, by introducing a risk enhancement factor based on the coefficient of variation, the evaluation model can automatically identify and strengthen the importance of highly volatile indicators. This overcomes the limitations of traditional static weight allocation models and solves the problem of strong subjectivity in weight setting during multi-source data fusion. Through the dynamic weighting adjustment mechanism, it can respond to changes in equipment status in real time, significantly improving the accuracy and timeliness of health assessments.

[0039] As a preferred approach, the weights of the primary indicators are dynamically calculated based on the comprehensive state characteristics of each primary indicator subsystem, including: S311. Calculate the weighted degradation degree of each first-level indicator subsystem. As shown in the following formula: ,in, Indicates the first Normalized degradation values ​​of each secondary indicator; S312. Calculate the average volatility of each first-level indicator subsystem. As shown in the following formula: ,in, Indicates the first Average volatility of each subsystem Representation Subsystem The number of secondary indicators included. Indicators coefficient of variation; S313, Based on weighted degradation and average volatility Calculate the comprehensive risk factor As shown in the following formula: ,in, Indicates the first Comprehensive risk factors for each subsystem represents the risk adjustment factor, and represents the weight preference for the degree of degradation, with a value range of [0, 1]; where when When the value approaches 1, it indicates a greater emphasis on the weighted degradation degree. (i.e., the comprehensive performance of the actual deterioration state of the indicators) is used to assess the overall risk of the subsystem, emphasizing the contribution of the current deterioration of the equipment to the risk; when When it approaches 0, more emphasis is placed on average volatility. The assessment uses a comprehensive evaluation of the degree of indicator volatility to focus on the potential risks arising from indicator fluctuations during equipment operation. Typically, both the degree of equipment degradation and indicator volatility are considered. Set to 0.5; S314. Based on comprehensive risk factors Dynamically allocate the weights of primary indicators As shown in the following formula: .

[0040] Understandably, in the dynamic weight adjustment scheme based on real-time fluctuation characteristics in this application, the coefficient of variation (CV) of each indicator is calculated using a sliding time window (e.g., 24 hours) to quantify the degree of fluctuation of the indicators; a risk enhancement factor (β) is designed (e.g., λ∈[1.2,1.5]), for example, when indicators such as current exhibit abnormal fluctuations, their weights will automatically increase by 20%-50%; finally, normalization processing is used to ensure the balance of the weight system and avoid excessive influence of a single indicator on the evaluation results.

[0041] In step S314 above, a comprehensive risk factor calculation model was constructed, enabling dynamic adjustment of the weights of primary indicators. Specifically, this is reflected in the following: First, a weighted degradation measure is used to quantify the overall health status of the subsystems, and normalization is used to ensure comparability between subsystems; second, a coefficient of variation is introduced to calculate the average volatility of the subsystems, effectively capturing potential risk signals; finally, a configurable risk adjustment factor γ is used to flexibly balance the influence weights of degradation degree and volatility.

[0042] In practical applications of the above method in this embodiment, when the vibration value of a bearing suddenly increases but does not exceed the threshold, the above method can dynamically adjust the weight ratio of the vibration index to promptly increase the weight ratio, enabling the health score to quickly respond to potential faults and detect anomalies earlier compared to the static weight method.

[0043] As a preferred embodiment, the step of calculating the membership degree of each secondary indicator to each preset health level based on the membership function includes: substituting the deterioration value of each secondary indicator into the following four membership functions: selecting a descending ridge distribution function to calculate its membership degree to the first health level; selecting an intermediate ridge distribution function to calculate its membership degree to the second and third health levels; and selecting an ascending ridge distribution function to calculate its membership degree to the fourth health level.

[0044] The membership degree of the distribution to the first health level is calculated using the descending ridge distribution function, as follows: ; The membership degree of the second health level is calculated using the intermediate ridge distribution function, as follows: The membership degree of the third health level is calculated using the intermediate ridge distribution function, as follows: The membership degree of the fourth health level is calculated using the ascending ridge distribution function, as shown in the following formula: Where, d i This indicates the degree of degradation. Health levels are defined in descending order of equipment condition: Level 1 corresponds to optimal operating condition; Level 2 corresponds to normal operating condition; Level 3 corresponds to a condition requiring attention; and Level 4 corresponds to a faulty or impending faulty condition. In simpler terms, optimal corresponds to excellent, normal to good, requiring attention to average, and faulty or impending faulty corresponds to a fault or warning. For more detailed information, please refer to... Figure 4 The diagram illustrates an example of a membership function image understanding scheme, where the horizontal axis represents the degradation value and the vertical axis represents the membership value.

[0045] For example, substituting the degradation value into the four membership functions, such as a vibration degradation of 0.4, its membership values ​​for optimal operating condition, normal operating condition, condition requiring attention, and fault are 0 (substituting into the first formula, 0.4 > 0.3), 1 (substituting into the second formula, = 0.4), 0 (substituting into the third formula, = 0.4), and 0 (substituting into the fourth formula, < 0.7), respectively. (0, 1, 0, 0) is a membership vector. Adding the membership vectors for temperature, voltage, and current forms the membership matrix of the operating condition indicators (i.e., ...). ).

[0046] As a preferred option, in step S5, the evaluation vector of each primary indicator is calculated. As shown in the following formula: ; in, This is the final weight vector for the secondary indicators. This represents the membership matrix of the i-th primary indicator subsystem (the membership vectors of multiple secondary indicators of the primary indicator constitute this membership matrix).

[0047] For example, for the operating status (first-level indicator), A1 = (a1, a2, a3, a4), where a1, a2, a3, and a4 represent the final weights of vibration, temperature, voltage, and current (second-level indicators), respectively. The weights of several second-level indicators form the indicator weight vector.

[0048] In step S5, the final evaluation vector B is calculated by combining the dynamically adjusted weights of the primary indicators, as shown in the following formula: ; In the formula This is the weight vector for the primary indicators. The evaluation matrix is ​​composed of the evaluation vectors of the primary indicators.

[0049] In step S6, the health level of the device is determined based on the principle of maximum membership, combined with the final evaluation vector.

[0050] For example, in the evaluation vector calculation stage: taking the operating state subsystem as an example, it includes four secondary indicators: vibration (weight 0.34), temperature (0.35), current (0.10), and voltage (0.21). First, the weight vector W is constructed. i =[0.34,0.35,0.10,0.21], and then calculate the membership matrix R based on the degradation values ​​of each index (vibration 0.4, temperature 0.5, etc.). i Through matrix multiplication The evaluation vector for this subsystem is obtained. For example, if the calculated value is [0, 0.8, 0.2, 0], it indicates that the subsystem's "good" level membership reaches 80%. In the comprehensive evaluation stage, the evaluation vectors of each subsystem are combined into a first-level evaluation matrix R, which may include the evaluation results of four subsystems: operating status, historical data, environmental parameters, and equipment attributes. Based on the dynamically adjusted first-level weights W=[0.36, 0.43, 0.12, 0.09], a comprehensive calculation is performed, resulting in a final evaluation vector such as [0.15, 0.45, 0.3, 0.1]. The "good" level membership is 45%, which is the maximum value. Therefore, the overall health of the equipment is judged to be "good".

[0051] Example 2 Based on the same inventive concept, a specific embodiment of this application proposes a device health evaluation system based on multi-dimensional data fusion, used to implement the device health evaluation method as described in Embodiment 1. The system includes: The data acquisition module is used to construct an evaluation system that includes multiple primary indicator subsystems, each of which contains several secondary indicators. The first calculation module is used to normalize the pre-collected secondary indicator data and calculate the degradation value of each secondary indicator. The second calculation module is used to calculate the initial combination weights of each secondary indicator, correct the initial combination weights based on the fluctuation characteristics of each secondary indicator, and obtain the final weights of the secondary indicators; and dynamically calculate the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem. The third calculation module is used to calculate the membership degree of each secondary indicator to each preset health level based on the membership function, and to construct the membership degree matrix; The fourth calculation module is used to calculate the evaluation vector of each primary indicator based on the final weights and membership matrix of the secondary indicators; and to calculate the final evaluation vector by combining the dynamically adjusted weights of the primary indicators. The health assessment module is used to determine the health level of the device based on the principle of maximum membership, by combining the final evaluation vector.

[0052] For example, multiple primary indicator subsystems include an operational status subsystem, a historical data subsystem, an environmental parameter subsystem, and an equipment attribute subsystem; the operational status subsystem includes secondary indicators such as vibration, temperature, current, and voltage; the historical data subsystem includes secondary indicators such as maintenance frequency, failure rate, and MTBF (Mean Time Between Failures); the environmental parameter subsystem includes secondary indicators such as ambient temperature, humidity, and dust concentration; and the equipment attribute subsystem includes secondary indicators such as design life and material durability.

[0053] Based on the above embodiments, the equipment health evaluation system proposed in this application achieves dynamic assessment through multi-module collaborative work. Its core working principle is as follows: The system first constructs a four-level evaluation system including operating status, historical data, environmental parameters, and equipment attributes through a data acquisition module, and collects data on 12 secondary indicators such as vibration and temperature in real time. The first calculation module normalizes the raw data using three types of degradation degree functions; for example, it uses a function where smaller values ​​are better to calculate the degradation degree for the temperature indicator. The second calculation module integrates the correlation coefficient method and the analytic hierarchy process (AHP) to determine the initial weights, and uses the coefficient of variation to construct a risk enhancement factor for dynamic correction; for example, it automatically increases the weight ratio of the vibration indicator when its fluctuations increase. The third calculation module substitutes the degradation degree value into the membership function. The fourth calculation module completes the evaluation from the indicator layer to the system layer through two-level fuzzy comprehensive evaluation. Finally, the health evaluation module outputs the health level based on the principle of maximum membership degree.

[0054] For specific limitations regarding the equipment health evaluation system based on multidimensional data fusion, please refer to the limitations of the equipment health evaluation method based on multidimensional data fusion mentioned above, which will not be repeated here. It should be noted that each module in the above evaluation system corresponds to steps S1 to S6 in implementing the above evaluation method. The instances and application scenarios implemented by multiple modules and their corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1 above.

[0055] In another embodiment of the present invention, an electronic device is also provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the device health evaluation method based on multidimensional data fusion as described in Embodiment 1.

[0056] In another embodiment of the present invention, a computer-readable storage medium is also provided, characterized in that it stores a computer program, which, when executed by a processor, implements the steps of the device health evaluation method based on multidimensional data fusion as described in Embodiment 1.

[0057] The specific implementation method and algorithm process of this method will be described in detail below with reference to actual processing routines.

[0058] 1. Calculate objective weights using secondary indicator data of equipment operating status over a period of time.

[0059] Table 1. Parameters of Equipment Operating Status over Five Days ; Based on the correlation coefficient method formula, the mean values ​​of vibration, temperature, current, voltage, and health score were calculated respectively, with results of 4, 40.2, 16.1, 231.1, and 81.6. Then, the correlation coefficients between each indicator and the health score were calculated. Taking vibration as an example, the numerator in the formula... The value is -6.34, and in the denominator... It is 0.71. The value is 141.2. Using the formula, the correlation coefficient between vibration and health is -0.63. Similarly, the correlation coefficients between temperature, current, and voltage and health are -0.68, -0.3, and 0.89, respectively. Then, the weights are calculated; the denominator in the weighting formula... With a value of 2.5, the objective weights for vibration, temperature, current, and voltage are calculated to be 0.252, 0.272, 0.12, and 0.356, respectively.

[0060] 2. Calculate subjective weights using the Analytic Hierarchy Process (AHP). Based on expert advice, a reasonable judgment matrix should be set as follows: ; Based on the above formula, the numerators of the vibration, temperature, current, and voltage indices can be calculated. The values ​​are 1.45, 2.06, 0.54, and 0.51 respectively, with the denominator being... The calculated subjective weights for vibration, temperature, current, and voltage are 0.32, 0.45, 0.12, and 0.12, respectively, with a value of 4.59. Since expert judgments may exhibit some inconsistency, a consistency check is required to assess the reliability of the judgment matrix. The largest eigenvalue of the judgment matrix is ​​obtained through matrix calculation. It is 3.95, calculated according to the formula. The value is -0.017. By looking up the table, we can see that when... At that time, the random consistency index The consistency ratio is 0.9. It is -0.019. The absolute values ​​are much less than 0.1, indicating good consistency. Therefore, the subjective weights of vibration, temperature, current, and voltage obtained through the analytic hierarchy process are 0.32, 0.45, 0.12, and 0.12, respectively.

[0061] 3. Calculate the portfolio weights: By weighting and taking into account both objective data and expert experience, the final combined weight is obtained and set to 0.6. Then, the combined weights of vibration, temperature, current and voltage are calculated using formulas to be 0.28, 0.34, 0.12 and 0.26, respectively.

[0062] 4. Combination weight adjustment: By calculating the coefficient of variation of the indicators, a risk enhancement factor was constructed, and the combined weights were corrected and normalized. After correction, the combined weights of vibration, temperature, current and voltage were 0.34, 0.35, 0.10 and 0.21, respectively.

[0063] 5. By calculating the weighted deterioration and average volatility of the subsystem, and then calculating the comprehensive risk factor, the first-level weights are dynamically adjusted and normalized. The weight of the first-level indicator of the operating status is adjusted to 0.36.

[0064] Table 2 Parameter Settings for Operational Status Indicator Deterioration Degree Function ; 6. Deterioration Calculation On a certain day, the operating data for vibration, temperature, current and voltage were 3.8 mm / s, 45℃, 16A and 230V, respectively. Using the corresponding degradation degree function in Table 2, the degradation degree corresponding to each index was calculated to be 0.4, 0.5, 0.3 and 0.2, respectively.

[0065] 7. Repeat the above process, performing the same calculations on historical data indicators (number of repairs, failure rate, and MTBF), environmental parameter indicators (ambient temperature, ambient humidity, and dust concentration), and equipment attribute indicators (design life and material durability). The results are shown in Table 3. It should be noted that the weights of the secondary indicators in the table are the results after correction, and the weights of the primary indicators are also dynamically adjusted.

[0066] Table 3. Operating data, degradation degree, and weight of each indicator. ; 8. Membership analysis The calculation and explanation are based on four secondary indicators of operating status (vibration, temperature, current, and voltage). Table 3 shows the degradation degree of each secondary indicator. Then, the membership function can be used to calculate the membership matrix of each indicator. The degradation degree of vibration is 0.4, and the membership degrees corresponding to very good, good, average, and faulty are 0, 1, 0, and 0, respectively. Similarly, the membership degrees of temperature are 0, 0.5, 0.5, and 0, the membership degrees of current are 0, 1, 0, and 0, and the membership degrees of voltage are 0.5, 0.5, 0, and 0.

[0067] Membership matrix of running state subsystem : ; And so on, the membership matrix of the historical data subsystem : ; Membership matrix of environmental parameter subsystem : ; Device Attribute Subsystem Membership Matrix : ; According to the formula ( For each subsystem, there is a weight vector of secondary indicators. (This is the membership matrix of each subsystem).

[0068] Running status subsystem: ; ; Historical data subsystem: ; ; Environmental parameter subsystem: ; ; Device Attribute Subsystem: ; ; Construct the primary indicator evaluation matrix and calculate the final evaluation vector: Primary indicator evaluation matrix: ; First-level indicator weight vector: ; Final evaluation vector: ; According to the principle of maximum membership, since 0.45 (the membership degree of the good level) is the maximum, the equipment health is "good".

[0069] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0070] In addition, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0071] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for evaluating equipment health based on multi-dimensional data fusion, applied in the field of industrial equipment health management, characterized in that, The method includes the following steps: S1. Construct an industrial equipment health evaluation system that includes four primary indicator subsystems: operating status, historical data, environmental parameters, and equipment attributes. Each primary indicator subsystem contains several secondary indicators that are directly related to the operating characteristics of industrial equipment. The secondary indicators of the operating status subsystem include vibration, temperature, current, and voltage. The secondary indicators of the historical data subsystem include maintenance frequency, failure rate, and mean time between failures. The secondary indicators of the environmental parameters subsystem include ambient temperature, humidity, and dust concentration. The secondary indicators of the equipment attributes subsystem include design life and material durability. S2. Normalize the pre-collected secondary indicator data and calculate the degradation value of each secondary indicator. S3. Calculate the initial combination weights of each secondary indicator, and correct the initial combination weights based on the fluctuation characteristics of each secondary indicator to obtain the final weights of the secondary indicators; dynamically calculate the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem. S4. Calculate the membership degree of each secondary indicator to each preset health level based on the membership function, and construct the membership degree matrix; S5. Based on the final weights and membership matrices of the secondary indicators, calculate the evaluation vector for each primary indicator. The final evaluation vector is calculated by combining the dynamically adjusted weights of the primary indicators; S6. Based on the final evaluation vector, determine the health level of the device according to the principle of maximum membership. In step S3, the calculation of the final weight of the secondary indicator includes: S301. The correlation coefficient method is used to calculate the correlation coefficient between the secondary indicators of each operating status and the health outcome indicators, and to obtain the objective weights. The correlation coefficient method uses the Pearson correlation coefficient, and the health outcome index is the expert score or the normalized value of the set performance parameter in the equipment's historical operation. S302. Construct a judgment matrix using the analytic hierarchy process (AHP) and calculate the subjective weights of each secondary indicator. ; S303, Comprehensive objective weights through weighted average and subjective weight To obtain the initial combined weights As shown in the following formula: ;in, This represents the objective weight of the j-th indicator. This represents the subjective weight of the j-th indicator. It is an adjustment parameter, with a value range of [0,1]. S304. Calculate the coefficient of variation of each secondary indicator within the preset time window. As shown in the following formula: ,in and This represents the standard deviation and mean of the secondary indicators within a preset time window. The volatility of a quantitative indicator is indicated by its value; the larger the value, the stronger the volatility of that secondary indicator. S305, Based on the aforementioned coefficient of variation Constructing risk enhancement factors As shown in the following formula: ;in, To amplify the upper limit of risk, The value range is set to ; S306. Utilizing the aforementioned risk enhancement factor For the initial combined weights Make corrections to obtain the corrected combined weights. As shown in the following formula: ; S307. Adjusted combined weights Normalization is performed to obtain the final weights of the secondary indicators. .

2. The device health evaluation method based on multi-dimensional data fusion according to claim 1, characterized in that, Before calculating the degradation value of each secondary indicator, the method further includes selecting a fitness function according to the data type of the secondary indicator, specifically including: For the first target indicator, the first degradation function is selected as follows: ； For the second target index, the second degradation function is selected as follows: ； For the third target indicator, the third degradation function is selected as follows: ； Among them, the smaller the value of the first target indicator, the better the equipment status; the value of the second target indicator is within the preset target range, indicating that the equipment status is optimal; and the larger the value of the third target indicator, the better the equipment status. The measured value of the indicator. The minimum value of the indicator. The maximum value of the indicator. This represents the lower limit of the target parameter range. This indicates the upper limit of the target parameter range.

3. The device health evaluation method based on multi-dimensional data fusion according to claim 1, characterized in that, The dynamic calculation of the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem includes: S311. Calculate the weighted degradation degree of each first-level indicator subsystem. As shown in the following formula: ,in, Indicates the first Normalized degradation values ​​of each secondary indicator; S312. Calculate the average volatility of each first-level indicator subsystem. As shown in the following formula: ,in, Indicates the first Average volatility of each subsystem Representation Subsystem The number of secondary indicators included. Indicators coefficient of variation; S313, Based on weighted degradation and average volatility Calculate the comprehensive risk factor As shown in the following formula: ,in, Indicates the first Comprehensive risk factors for each subsystem , which is a risk adjustment factor, representing the weight preference for the degree of degradation, and its value ranges from [0, 1]. S314. Based on comprehensive risk factors Dynamically allocate the weights of primary indicators As shown in the following formula: .

4. The equipment health evaluation method based on multi-dimensional data fusion according to claim 1, characterized in that, The calculation of the membership degree of each secondary indicator to each preset health level based on the membership function includes: substituting the deterioration value of each secondary indicator into the following four membership functions: A descending ridge distribution function is selected to calculate its membership degree to the first health level; an intermediate ridge distribution function is selected to calculate its membership degree to the second and third health levels; and an ascending ridge distribution function is selected to calculate its membership degree to the fourth health level. The health levels are defined in order of the degree of equipment condition: the first health level corresponds to the equipment being in the best operating condition; the second health level corresponds to the equipment being in the normal operating condition; the third health level corresponds to the equipment being in a state requiring attention; and the fourth health level corresponds to the equipment being in a faulty or imminent faulty state.

5. The equipment health evaluation method based on multi-dimensional data fusion according to claim 4, characterized in that, The evaluation vector for each primary indicator is calculated. As shown in the following formula: ； in, This is the final weight vector for the secondary indicators. Let represent the membership matrix of the i-th first-level index subsystem.

6. The device health evaluation method based on multi-dimensional data fusion according to claim 5, characterized in that, The final evaluation vector B is calculated by combining the dynamically adjusted weights of the primary indicators, as shown in the following formula: ； In the formula This is the weight vector for the primary indicators. The evaluation matrix is ​​composed of the evaluation vectors of the primary indicators.

7. A multi-dimensional data fusion-based equipment health assessment system, applied in the field of industrial equipment health management, for implementing the equipment health assessment method as described in any one of claims 1-6, characterized in that, The system includes: The data acquisition module is used to construct an industrial equipment health evaluation system that includes four primary indicator subsystems: operating status, historical data, environmental parameters, and equipment attributes. Each primary indicator subsystem contains several secondary indicators that are directly related to the operating characteristics of industrial equipment. The secondary indicators of the operating status subsystem include vibration, temperature, current, and voltage. The secondary indicators of the historical data subsystem include maintenance frequency, failure rate, and mean time between failures. The secondary indicators of the environmental parameters subsystem include ambient temperature, humidity, and dust concentration. The secondary indicators of the equipment attributes subsystem include design life and material durability. The first calculation module is used to normalize the pre-collected secondary indicator data and calculate the degradation value of each secondary indicator. The second calculation module is used to calculate the initial combination weights of each secondary indicator, correct the initial combination weights based on the fluctuation characteristics of each secondary indicator, and obtain the final weights of the secondary indicators; and dynamically calculate the weights of the primary indicators based on the comprehensive state characteristics of each primary indicator subsystem. The third calculation module is used to calculate the membership degree of each secondary indicator to each preset health level based on the membership function, and to construct the membership degree matrix; The fourth calculation module is used to calculate the evaluation vector of each primary indicator based on the final weights and membership matrix of the secondary indicators; and to calculate the final evaluation vector by combining the dynamically adjusted weights of the primary indicators. The health assessment module is used to determine the health level of the device based on the principle of maximum membership, by combining the final evaluation vector.

8. An electronic device, characterized in that, The device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the device health evaluation method based on multidimensional data fusion as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The device contains a computer program that, when executed by a processor, implements the steps of the device health assessment method based on multidimensional data fusion as described in any one of claims 1-6.

Images