Mechanical equipment state monitoring method and device based on dynamic heterogeneous period weighting index

Abstract

The application discloses a kind of mechanical equipment condition monitoring method and device based on dynamic heterocyclic weighting index, the method comprises: the current vibration data of equipment is collected;Based on current vibration data, according to the weighted superposition of the waveform factor corresponding to vibration data in two adjacent time periods, obtain early warning index;If early warning index is greater than preset early warning threshold, then send early warning signal;Wherein, the calculation parameter of early warning index is determined by the historical vibration data of equipment and preset fitness function.The application can send early warning in time in the early stage of equipment failure, with strong adaptability and high accuracy.

Description

Technical Field

[0001] This invention relates to the field of mechanical equipment condition monitoring technology, and in particular to a method and device for mechanical equipment condition monitoring based on dynamic inter-period weighted index. Background Technology

[0002] With the development of technology, mechanical equipment is widely used in various fields of modern life, work, and production. Currently, mechanical equipment is developing towards high precision and sophistication, and its operating environment is becoming increasingly complex and variable. Continuous operation, prolonged periods of operation, frequent changes in working conditions, and harsh operating environments can all affect the health of mechanical equipment, thereby reducing its operating efficiency. Therefore, it is necessary to monitor the condition of mechanical equipment to promptly identify and resolve equipment problems.

[0003] Existing methods for monitoring the condition of mechanical equipment are mainly divided into two categories: direct observation and information monitoring. Direct observation relies on the experience and judgment of technicians, resulting in high subjectivity and poor accuracy, making comprehensive and accurate real-time monitoring difficult. Information monitoring methods mainly include vibration analysis, oil sample analysis, temperature monitoring, and acoustic emission methods. Among these, vibration analysis is susceptible to noise interference and struggles to accurately identify early faults during low-speed or intermittent operation; oil sample analysis suffers from time lag and cannot achieve real-time monitoring; temperature monitoring is insensitive to minor faults and exhibits a lag in response; while acoustic emission can detect and monitor microcracks, it is easily affected by environmental noise and is costly. Existing methods suffer from poor real-time performance, accuracy, and adaptability. Summary of the Invention

[0004] To address the above technical problems, this invention provides a method and device for monitoring the condition of mechanical equipment based on dynamic inter-period weighted indices. This method can issue timely warnings in the early stages of equipment failure and has strong adaptability and high accuracy.

[0005] This invention provides a method for monitoring the condition of mechanical equipment based on a dynamic, inter-period weighted index, comprising: collecting current vibration data of the equipment;

[0006] Based on the current vibration data, an early warning index is calculated by weighting and superimposing the waveform factors corresponding to the vibration data in two adjacent time periods.

[0007] If the warning indicator is greater than the preset warning threshold, a warning signal is issued;

[0008] The calculation parameters of the early warning index are determined by the historical vibration data of the equipment and a preset fitness function.

[0009] As an improvement to the above scheme, the calculation parameters of the early warning indicator include: a first time parameter, a second time parameter, and a weighting coefficient of the waveform factor;

[0010] The method for determining the calculation parameters includes:

[0011] Acquire historical vibration data of the device;

[0012] Based on the historical vibration data and the preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters.

[0013] As an improvement to the above scheme, the early warning index is calculated based on the current vibration data and the weighted superposition of waveform factors corresponding to vibration data in two adjacent time periods, including:

[0014] Based on the current vibration data, the warning index for the current moment is calculated according to the following formula:

[0015] F(t1,t2,t,α)=α×f(t2,t1)+(1-α)×f(t1,t)

[0016] Where F(t1,t2,t,α) is the warning index at the current time t, t1 is the first time parameter, t2 is the second time parameter, t≥t2>t1>1, and α is the weighting coefficient of the waveform factor; f(t2,t1) and f(t1,t) are the waveform factors of the vibration data collected in the first time period [t-t2,t-t1] and the second time period [t-t1,t], respectively.

[0017] The formula for calculating the waveform factor is:

[0018]

[0019] Where x = {x1, x2, ..., x} N}, where x represents the vibration data collected within a set time period; N is the total number of data points within the time period.

[0020] As an improvement to the above scheme, the range of values ​​for the weighting coefficient α is:

[0021] As an improvement to the above scheme, the formula for the fitness function Zf is:

[0022]

[0023] Where x represents the time of the first warning; σ is the average warning time; σ is the standard deviation of the warning time; y is the warning index value at the time of the first warning; the warning time is determined according to the warning index.

[0024] As an improvement to the above scheme, the optimization of the calculation parameters of the early warning index with the goal of minimizing the fitness function specifically includes:

[0025] With the goal of minimizing the fitness function, the calculation parameters of the early warning index are optimized using the Hippo optimization algorithm.

[0026] As an improvement to the above scheme, the hippo optimization algorithm specifically includes:

[0027] Initialize the positions of several individual hippos, where the position of each individual hippo is composed of the weighting coefficients of the first time parameter, the second time parameter, and the waveform factor;

[0028] By simulating the behavior of a hippopotamus searching for food in water, the location was adjusted to perform a global search.

[0029] By simulating the behavior of hippos defending against predators, the location was dynamically adjusted to enhance population diversity;

[0030] By simulating the behavior of a hippopotamus escaping a predator, a local search was conducted, and the location was fine-tuned.

[0031] When the number of iterations reaches the set maximum number of iterations, the globally optimal parameter combination is output.

[0032] As an improvement to the above scheme, the warning threshold is determined according to the Raida criterion.

[0033] This invention also provides a mechanical equipment condition monitoring device based on a dynamic, inter-period weighted index, comprising:

[0034] The data acquisition module is used to collect the current vibration data of the equipment;

[0035] The indicator calculation module is used to calculate the early warning indicator based on the current vibration data and the weighted superposition of the waveform factors corresponding to the vibration data in two adjacent time periods.

[0036] The early warning module is used to issue an early warning signal if the early warning indicator is greater than a preset early warning threshold.

[0037] The calculation parameters of the early warning index are determined by the historical vibration data of the equipment and a preset fitness function.

[0038] Furthermore, the calculation parameters of the early warning indicator include: a first time parameter, a second time parameter, and a weighting coefficient of the waveform factor;

[0039] The device is also used for:

[0040] Acquire historical vibration data of the device;

[0041] Based on the historical vibration data and the preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters.

[0042] Compared to existing technologies, the beneficial effects of the mechanical equipment condition monitoring method and device based on dynamic inter-period weighted index provided by this invention are as follows: By calculating an early warning index based on the current vibration data and the weighted superposition of waveform factors corresponding to vibration data in two adjacent time periods, an early warning signal is issued when the early warning index exceeds a preset early warning threshold, enabling timely early warning at the early stage of equipment failure. By designing a fitness function and employing an improved optimization algorithm for parameter optimization, the calculation parameters of the early warning index can be adjusted according to different equipment and working conditions, improving the adaptability and accuracy of equipment condition monitoring. This invention, by using a dynamic inter-period weighted index for mechanical equipment condition monitoring, effectively improves the operational safety and reliability of the equipment. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating a mechanical equipment condition monitoring method based on dynamic inter-period weighted indices provided in an embodiment of the present invention.

[0044] Figure 2 This is a time series diagram of the early warning indicators of the training set provided in this embodiment of the invention;

[0045] Figure 3 This is a graph showing the optimization effect of the fitness function provided in an embodiment of the present invention;

[0046] Figure 4 This is a time-domain graph corresponding to the training set generated according to the national standard method, provided in an embodiment of the present invention.

[0047] Figure 5 This is a time-domain graph corresponding to the training set generated based on the dynamic inter-period weighted index, provided in an embodiment of the present invention.

[0048] Figure 6 This is a schematic diagram of the structure of a mechanical equipment condition monitoring device based on a dynamic inter-period weighted index provided in an embodiment of the present invention. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0050] Please see Figure 1 , Figure 1 This is a flowchart illustrating a mechanical equipment condition monitoring method based on a dynamic inter-period weighted index, provided by an embodiment of the present invention. The mechanical equipment condition monitoring method based on a dynamic inter-period weighted index includes: S1: collecting current vibration data of the equipment;

[0051] S2: Based on the current vibration data, the early warning index is calculated by weighting and superimposing the waveform factors corresponding to the vibration data in two adjacent time periods;

[0052] S3: If the warning indicator is greater than the preset warning threshold, a warning signal is issued;

[0053] The calculation parameters of the early warning index are determined by the historical vibration data of the equipment and a preset fitness function.

[0054] Specifically, this invention incorporates a dynamic, inter-period weighted index, also known as a warning index. Furthermore, by collecting real-time vibration acceleration data of the mechanical equipment and calculating the current warning index value based on this data, an abnormal vibration is detected when the warning index value exceeds a preset warning threshold. This triggers a warning signal to prompt technicians to maintain the equipment promptly. The calculation parameters for the warning index are adjustable, allowing for adaptive adjustments based on different equipment or operating conditions.

[0055] As one optional embodiment, the calculation parameters of the early warning indicator include: a first time parameter, a second time parameter, and a weighting coefficient of the waveform factor.

[0056] As one optional embodiment, the step of calculating the early warning index based on the current vibration data and the weighted superposition of waveform factors corresponding to vibration data in two adjacent time periods includes:

[0057] Based on the current vibration data, the warning index for the current moment is calculated according to the following formula:

[0058] F(t1,t2,t,α)=α×f(t2,t1)+(1-α)×f(t1,t)

[0059] Where F(t1, t2, t, α) is the warning index at the current time t, t1 is the first time parameter, t2 is the second time parameter, t≥t2>t1>1, and α is the weighting coefficient of the waveform factor; f(t2, t1) and f(t1, t) are the waveform factors of the vibration data collected in the first time period [t-t2, t-t1] and the second time period [t-t1, t], respectively.

[0060] The formula for calculating the waveform factor is:

[0061]

[0062] Where x = {x1, x2, ..., x} N}, where x represents the vibration data collected within a set time period; N is the total number of data points within the time period.

[0063] As one optional embodiment, the range of values ​​for the weighting coefficient α is:

[0064] By setting the range of values ​​for the weighting coefficient α, sufficient degrees of freedom can be provided to adjust the weighting strategy, and when using the fitness function for optimization, the convergence process can be accelerated and unnecessary calculations can be reduced, avoiding redundancy in the search process due to an excessively large parameter space.

[0065] As one optional embodiment, the method for determining the calculation parameters includes:

[0066] Acquire historical vibration data of the device;

[0067] Based on the historical vibration data and the preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters.

[0068] Specifically, based on the historical vibration data of the mechanical equipment, the early warning index corresponding to the historical vibration data is calculated, a fitness function is set, and then the parameters are optimized by combining the early warning index and the fitness function to obtain the optimized early warning index calculation formula, so as to carry out real-time monitoring of equipment status and fault early warning.

[0069] As one optional embodiment, the fitness function Z f The formula is:

[0070]

[0071] Where x represents the time of the first warning; σ is the average warning time; σ is the standard deviation of the warning time; y is the warning index value at the time of the first warning; the warning time is determined according to the warning index.

[0072] Specifically, the first warning time is the time when a potential fault is detected and a warning is issued; the average warning time is the average of all warning times recorded in a series of tests (i.e., historical vibration data); and the standard deviation of the warning time is used to measure the dispersion of the warning time in a series of tests.

[0073] The first part of the fitness function formula: This is used to amplify situations where the first warning time is later than the average warning time, thereby increasing the value of the fitness function. When the first warning time is significantly longer than the average warning time, exponential growth can impose a heavy penalty, accelerating the growth of the fitness function value and prompting the optimization function to avoid this situation.

[0074] The second part of the fitness function formula, -log(y+1), encourages high alert values. When y is large, the growth rate of the logarithmic function slows down, but the overall growth remains negative, resulting in a smaller fitness function value. This design encourages issuing alerts when more serious anomalies are detected, thus enabling early identification of potential major problems. y+1 is used to prevent the logarithmic function from failing to calculate when y is zero, while ensuring that the logarithmic input is positive.

[0075] As one optional embodiment, optimizing the calculation parameters of the early warning index with the objective of minimizing the fitness function specifically includes:

[0076] With the goal of minimizing the fitness function, the calculation parameters of the early warning index are optimized using the Hippo optimization algorithm.

[0077] As one optional embodiment, the hippo optimization algorithm specifically includes:

[0078] Initialize the positions of several individual hippos, where the position of each individual hippo is composed of the weighting coefficients of the first time parameter, the second time parameter, and the waveform factor;

[0079] By simulating the behavior of a hippopotamus searching for food in water, the location was adjusted to perform a global search.

[0080] By simulating the behavior of hippos defending against predators, the location was dynamically adjusted to enhance population diversity;

[0081] By simulating the behavior of a hippopotamus escaping a predator, a local search was conducted, and the location was fine-tuned.

[0082] When the number of iterations reaches the set maximum number of iterations, the globally optimal parameter combination is output.

[0083] Specifically, the Hippopotamus Optimization (HO) algorithm is inspired by the behavior of hippos in their natural environment. It simulates their activity patterns in rivers or ponds and uses global and local search strategies to find the optimal solution. The Hippopotamus Optimization algorithm mainly includes the following stages:

[0084] Initialization: During the initialization phase, the HO algorithm randomly generates a certain number of individuals (i.e., search agents) and assigns them positions. The parameters t1, t2, and α of these individuals are initialized within a range. For example, t1 might be between 1 and 5, t2 between 2 and 6, and α between 1 / 10 and 1 (depending on the set upper and lower boundaries). The parameters of each individual may be different, representing different candidate solutions. The positions of the randomly generated hippopotamus individuals serve as the initial population.

[0085] Exploration Phase: In the exploration phase, the algorithm simulates the behavior of a hippopotamus searching for food in the water, performing a global search process. The algorithm explores the entire search space. For optimization tasks, the exploration phase helps avoid getting trapped in local optima. During this phase, the algorithm randomly adjusts the parameters, i.e., varies t1, t2, and α, to fully explore the search space. For example, an individual hippopotamus moves to a new area guided by "food," which represents the exploration of parameters t1, t2, and α within the search space. The goal of this phase is to ensure that the initial selection of parameters covers the widest possible space, finding potential optimal solutions.

[0086] Defense Phase: In the defense phase, the algorithm simulates the behavior of a hippopotamus defending against predators. Here, the algorithm enhances population diversity and global search capabilities, introducing more randomness to prevent the search from being confined to a specific region. Specifically, this phase improves the algorithm's robustness by dynamically adjusting the parameters t1, t2, and α of individuals. For example, some parameters may be increased or decreased in some individuals, allowing the algorithm to remain stable when facing a more complex search space, thus avoiding premature convergence. The defense phase is a crucial part of the algorithm's strategy to avoid getting trapped in local optima.

[0087] Escape Phase: In the escape phase, the algorithm simulates the behavior of a hippo fleeing a predator, fine-tuning parameters and performing a local search to ensure a more accurate solution is found. During this phase, by fine-tuning t1, t2, and α, a more precise optimal parameter combination is found. By reducing the magnitude of parameter changes, the escape phase allows the algorithm to gradually approach the optimal solution.

[0088] Through the synergistic effect of the above stages, the Hippo optimization algorithm can efficiently balance exploration and utilization in the search space, thereby finding the global optimal solution to the optimization problem.

[0089] As one optional embodiment, the warning threshold is determined according to the Raida criterion.

[0090] Specifically, the Laida criterion, also known as the 3σ criterion, is used to determine the warning threshold. The warning threshold is obtained by calculating the sum of the mean and standard deviation of historical vibration data. It should be noted that the warning threshold obtained from different datasets will vary, but this difference is very small, and this minor difference can be further adjusted and optimized by optimizing the fitness function.

[0091] In a specific example, taking bearing condition detection as an example, two PCB 352C33 unidirectional accelerometers were used to acquire vibration data throughout the bearing's entire life cycle. These sensors were fixed to the bearing in the horizontal and vertical directions using magnetic mounts. Signal acquisition was performed using a DT9837 portable dynamic signal acquisition unit, with a sampling frequency of 25.6kHz, a sampling interval of 1 minute, and each sampling lasting 1.28 seconds. The acquired bearing vibration data is shown in Table 1.

[0092] Table 1

[0093] bearings Damaged location Total number of samples life Bearing 1-2 Outer ring 161 2h41min Bearing 1-3 Outer ring 158 2h38min Bearing2-2 Outer ring 161 2h41min Bearing 3-4 Inner circle 1515 25h15min

[0094] Vibration data from bearings 1-3 were used as the training set, and vibration data from the other three bearings were used as the test set. Initial calculation parameters were set according to the pre-defined warning indicators: t1 = 2, t2 = 4, and... Calculate the early warning indicators corresponding to the training set, and the calculation results are as follows: Figure 2 As shown, from Figure 2 As can be seen, the indicator curve is relatively stable for the first 50 minutes, but begins to fluctuate sharply after 55 minutes. Under this indicator, the time when the fluctuation begins represents the time when the fault begins. To eliminate the subjectivity of manually defining the fluctuation location, a unified early warning standard needs to be established to locate the fluctuation point in order to achieve the effect of monitoring the equipment status. To determine the fluctuation location, the Raida criterion is adopted to set the early warning threshold. By calculating the sum of the mean and standard deviation of the training set data, the alarm threshold is obtained as 1.3.

[0095] Furthermore, based on the fitness function, the hippo optimization algorithm is used for iterative parameter optimization, with the number of hippo individuals set to 50 and the maximum number of iterations set to 100. The optimization effect of the fitness function is as follows: Figure 3 As shown, by applying the Hippo Optimization Algorithm, the algorithm successfully avoided getting trapped in local optima and reached the optimal fitness value in the sixth generation. During the optimization process, the Hippo Optimization Algorithm effectively enhanced the early warning indicators, and the final optimized parameters obtained were: t1 = 1, t2 = 5, and...

[0096] To verify the effectiveness of the dynamic inter-period weighted index (i.e., early warning index) in the embodiments of the present invention, please refer to... Figure 4 and Figure 5 , Figure 4 This is a time-domain graph corresponding to the training set generated according to the national standard method, provided in an embodiment of the present invention. Figure 5 This is a time-domain graph of the training set generated based on the dynamic inter-period weighted index, provided in an embodiment of the present invention. It can be seen that after re-representing the training set data using the dynamic inter-period weighted index, the obtained warning time is 116 minutes. According to relevant national standards, the prescribed warning time is 150 minutes. Compared to the standard value, the optimized warning time is 34 minutes earlier, fully demonstrating the significant effect of the dynamic inter-period weighted index in optimizing the warning time.

[0097] The test set was used for verification, as shown in Table 2, which demonstrates the effectiveness of the test set. It can be seen that the early warning time obtained based on the dynamic inter-period weighted index of this invention can be as early as 27 minutes earlier than traditional methods, thus enabling earlier warnings of potential risks. This invention not only improves the timeliness of early warnings but also effectively enhances the response capability of the equipment monitoring system, providing more sufficient time to respond to potential risks.

[0098]

[0099] The early warning indicator in this embodiment of the invention comprehensively evaluates the overall characteristics of vibration data by combining time-series data from different time periods. The two time periods in the formula respectively cover earlier and more recent time-series data, enabling the indicator to simultaneously reflect long-term trends and recent changes. This helps to fully understand the operating status of the equipment and identify potential anomalies and faults. The indicator can quantify the degree of extreme fluctuations in the time-series data. Extreme fluctuations are often early signs of equipment failure, such as bearing wear, breakage, or poor lubrication. After combining data from different time periods, the weighted waveform factor can more sensitively capture these abnormal fluctuations, improving the accuracy and timeliness of fault monitoring and detection. The weighting coefficients in the indicator control the contribution ratio of historical and current data to the final waveform factor, allowing the dynamic inter-period weighted indicator to retain useful information from historical data while quickly responding to changes in current data. By optimizing and adjusting the indicator parameters through an optimization algorithm, the influence of historical and current data can be dynamically balanced according to the specific operating conditions of the equipment and the monitoring objectives. This allows the dynamic inter-period weighted indicator to be widely applied to various mechanical equipment and their variable operating environments, significantly improving the accuracy and adaptability of monitoring.

[0100] The embodiments of this invention can effectively monitor equipment vibration fluctuations and promptly detect potential faults. It is not only applicable to specific types of mechanical equipment but can also be widely used for vibration state analysis of various types of machinery. Through precise analysis of equipment vibration signals, this invention can achieve early fault warning during equipment operation, effectively improving the operational safety and reliability of the equipment.

[0101] Accordingly, the present invention also provides a mechanical equipment condition monitoring device based on dynamic inter-period weighted index, which can realize all the processes of the mechanical equipment condition monitoring method based on dynamic inter-period weighted index in the above embodiments.

[0102] Please see Figure 6 , Figure 6 This is a schematic diagram of a mechanical equipment condition monitoring device based on a dynamic inter-period weighted index, provided in an embodiment of the present invention. The mechanical equipment condition monitoring device based on the dynamic inter-period weighted index includes: a data acquisition module 601, used to acquire current vibration data of the equipment;

[0103] The indicator calculation module 602 is used to calculate the early warning indicator based on the current vibration data and the weighted superposition of the waveform factors corresponding to the vibration data in two adjacent time periods.

[0104] The early warning module 603 is used to issue an early warning signal if the early warning indicator is greater than a preset early warning threshold.

[0105] The calculation parameters of the early warning index are determined by the historical vibration data of the equipment and a preset fitness function.

[0106] Preferably, the calculation parameters of the early warning indicator include: a first time parameter, a second time parameter, and a weighting coefficient of the waveform factor;

[0107] The device is also used for:

[0108] Acquire historical vibration data of the device;

[0109] Based on the historical vibration data and the preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters.

[0110] Preferably, the step of calculating the early warning index based on the current vibration data and the weighted superposition of waveform factors corresponding to vibration data in two adjacent time periods includes:

[0111] Based on the current vibration data, the warning index for the current moment is calculated according to the following formula:

[0112] F(t1,t2,t,α)=α×f(t2,t1)+(1-α)×f(t1,t)

[0113] Where F(t1, t2, t, α) is the warning index at the current time t, t1 is the first time parameter, t2 is the second time parameter, t≥t2>t1>1, and α is the weighting coefficient of the waveform factor; f(t2, t1) and f(t1, t) are the waveform factors of the vibration data collected in the first time period [t-t2, t-t1] and the second time period [t-t1, t], respectively.

[0114] The formula for calculating the waveform factor is:

[0115]

[0116] Where x = {x1, x2, ..., x} N}, where x represents the vibration data collected within a set time period; N is the total number of data points within the time period.

[0117] Preferably, the range of values ​​for the weighting coefficient α is:

[0118] Preferably, the fitness function Z f The formula is:

[0119]

[0120] Where x represents the time of the first warning; σ is the average warning time; σ is the standard deviation of the warning time; y is the warning index value at the time of the first warning; the warning time is determined according to the warning index.

[0121] Preferably, optimizing the calculation parameters of the early warning index with the objective of minimizing the fitness function specifically includes:

[0122] With the goal of minimizing the fitness function, the calculation parameters of the early warning index are optimized using the Hippo optimization algorithm.

[0123] Preferably, the hippo optimization algorithm specifically includes:

[0124] Initialize the positions of several individual hippos, where the position of each individual hippo is composed of the weighting coefficients of the first time parameter, the second time parameter, and the waveform factor;

[0125] By simulating the behavior of a hippopotamus searching for food in water, the location was adjusted to perform a global search.

[0126] By simulating the behavior of hippos defending against predators, the location was dynamically adjusted to enhance population diversity;

[0127] By simulating the behavior of a hippopotamus escaping a predator, a local search was conducted, and the location was fine-tuned.

[0128] When the number of iterations reaches the set maximum number of iterations, the globally optimal parameter combination is output.

[0129] Preferably, the warning threshold is determined according to the Raida criterion.

[0130] In specific implementation, the working principle, control process and technical effects of the mechanical equipment condition monitoring device based on dynamic inter-period weighted index provided in the embodiments of the present invention are the same as those of the mechanical equipment condition monitoring method based on dynamic inter-period weighted index in the above embodiments, and will not be repeated here.

[0131] This invention provides a method and apparatus for monitoring the condition of mechanical equipment based on a dynamic, inter-period weighted index. Its advantages include: calculating an early warning index based on current vibration data and the weighted superposition of waveform factors corresponding to vibration data from two adjacent time periods; issuing an early warning signal when the index exceeds a preset threshold, enabling timely warnings at the early stages of equipment failure; and improving the adaptability and accuracy of equipment condition monitoring by designing a fitness function and employing an improved optimization algorithm to adjust the calculation parameters of the early warning index according to different equipment and operating conditions. This invention, by using a dynamic, inter-period weighted index for mechanical equipment condition monitoring, effectively improves the operational safety and reliability of the equipment.

[0132] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for monitoring the condition of mechanical equipment based on a dynamic, inter-period weighted index, characterized in that, include: Acquire current vibration data of the equipment; Based on the current vibration data, an early warning index is calculated by weighting and superimposing the waveform factors corresponding to the vibration data in two adjacent time periods. If the warning indicator is greater than the preset warning threshold, a warning signal is issued; Specifically, based on historical vibration data and a preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters. The early warning index is calculated based on the current vibration data and the weighted superposition of waveform factors corresponding to vibration data in two adjacent time periods, including: Based on the current vibration data, the warning index for the current moment is calculated according to the following formula: in, For the current moment Early warning indicators For the first time parameter, This is the second time parameter. , These are the weighting coefficients for the waveform factors; and The first time period Inner and Second Time Periods Waveform factor of internally acquired vibration data; The fitness function The formula is: in, This is the time of the first warning; This is the average warning time; The standard deviation of the warning time; The warning indicator value is the value of the warning indicator at the time of the first warning; the warning time is determined based on the warning indicator.

2. The mechanical equipment condition monitoring method based on dynamic inter-period weighted index as described in claim 1, characterized in that, The calculation parameters of the early warning indicator include: a first time parameter, a second time parameter, and a weighting coefficient of the waveform factor; The method for determining the calculation parameters includes: Acquire historical vibration data of the device; Based on the historical vibration data and the preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters.

3. The mechanical equipment condition monitoring method based on dynamic inter-period weighted index as described in claim 2, characterized in that, The formula for calculating the waveform factor is: in, , This indicates vibration data collected within a set time period; This represents the total number of data points within the time period.

4. The mechanical equipment condition monitoring method based on dynamic inter-period weighted index as described in claim 3, characterized in that, The weighting coefficients The range of values ​​for is: , .

5. The mechanical equipment condition monitoring method based on dynamic inter-period weighted index as described in claim 2, characterized in that, The optimization of the calculation parameters of the early warning index with the goal of minimizing the fitness function specifically includes: With the goal of minimizing the fitness function, the calculation parameters of the early warning index are optimized using the Hippo optimization algorithm.

6. The mechanical equipment condition monitoring method based on dynamic inter-period weighted index as described in claim 5, characterized in that, The hippo optimization algorithm specifically includes: Initialize the positions of several individual hippos, where the position of each individual hippo is composed of the weighting coefficients of the first time parameter, the second time parameter, and the waveform factor; By simulating the behavior of a hippopotamus searching for food in water, the location was adjusted to perform a global search. By simulating the behavior of hippos defending against predators, the location was dynamically adjusted to enhance population diversity; By simulating the behavior of a hippopotamus escaping a predator, a local search was conducted, and the location was fine-tuned. When the number of iterations reaches the set maximum number of iterations, the globally optimal parameter combination is output.

7. The method for monitoring the condition of mechanical equipment based on dynamic inter-period weighted indices as described in claim 1, characterized in that, The warning threshold is determined according to the Raida criterion.

8. A mechanical equipment condition monitoring device based on a dynamic, inter-period weighted index, characterized in that, include: The data acquisition module is used to collect the current vibration data of the equipment; The indicator calculation module is used to calculate the early warning indicator based on the current vibration data and the weighted superposition of the waveform factors corresponding to the vibration data in two adjacent time periods. The early warning module is used to issue an early warning signal if the early warning indicator is greater than a preset early warning threshold. Specifically, based on historical vibration data and a preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters. The early warning index is calculated based on the current vibration data and the weighted superposition of waveform factors corresponding to vibration data in two adjacent time periods, including: Based on the current vibration data, the warning index for the current moment is calculated according to the following formula: in, For the current moment Early warning indicators For the first time parameter, This is the second time parameter. , These are the weighting coefficients for the waveform factors; and The first time period Inner and Second Time Periods Waveform factor of internally acquired vibration data; The fitness function The formula is: in, This is the time of the first warning; This is the average warning time; The standard deviation of the warning time; The warning indicator value is the value of the warning indicator at the time of the first warning; the warning time is determined based on the warning indicator.

9. The mechanical equipment condition monitoring device based on dynamic inter-period weighted index as described in claim 8, characterized in that, The calculation parameters of the early warning indicator include: a first time parameter, a second time parameter, and a weighting coefficient of the waveform factor; The device is also used for: Acquire historical vibration data of the device; Based on the historical vibration data and the preset fitness function, the calculation parameters of the early warning index are optimized with the goal of minimizing the fitness function, resulting in optimized calculation parameters.

Images