Method and system for monitoring running state of wet washing machine

Abstract

The invention relates to the technical field of monitoring, in particular to a wet washing machine operation state monitoring method and system, and the method comprises the steps: obtaining a plurality of feature parameters of a wet washing machine and the state type of the wet washing machine, and calculating the mutual information entropy of each feature parameter and the state type; classifying the feature parameters by using a preset XGboost model to obtain a state category of the wet cleaning machine, and calculating an importance coefficient and classification accuracy of each decision tree in the XGboost model; and classifying the characteristic parameters obtained in real time by using the decision tree to obtain classification results, and performing decision fusion on the classification results through a voting method to obtain a final decision of the state category of the wet washing machine. According to the method, the working state of the wet cleaning machine is comprehensively analyzed by utilizing the plurality of characteristic parameters, and compared with a result obtained by analyzing a single factor, the dependence of the analysis result on the single factor is reduced, and the accuracy and reliability of a classification result are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of monitoring, and particularly to a method and system for monitoring the operating state of a wet washing machine. Background Art

[0002] A wet washing machine is a device used for cleaning and processing different types of items, and is usually applied in industries such as textiles, industrial parts, and electronic devices. The working principle of a wet washing machine is similar to that of a traditional washing machine, but its design and application scope are more diverse, and it can adapt to more professional cleaning requirements. When a wet washing machine for textiles is working, it is necessary to monitor its working state, so as to be able to understand the working state of the wet washing machine in real time and avoid losses to the production process caused by mechanical failures of the wet washing machine.

[0003] The Chinese patent application document with the publication number CN115574930A discloses a method for monitoring the working state of mechanical equipment. The method includes: obtaining the characteristic parameters of the equipment operation sound; storing the characteristic parameters as reference values in a specified memory; when the equipment is working, collecting the equipment operation sound in real time through a sound collector, and then extracting the characteristic parameters of the collected sound and comparing them with the reference values of the characteristic parameters; setting a decision threshold, comparing the comparison result with the decision threshold, and then outputting the detection result.

[0004] It can be seen from the above solution that when monitoring the working state of a wet washing machine, the working state of the wet washing machine can be judged by using the sound characteristic parameters. However, when judging the working state of the equipment by relying on the sound characteristic parameters and monitoring with a single factor, the result is also greatly affected by the single factor, resulting in poor stability of the monitoring result, increasing the possibility of misjudgment, and the accuracy of the monitoring result is relatively low. Summary of the Invention

[0005] In order to solve the problem that the accuracy of the monitoring result is relatively low due to judging the working state of the wet washing machine by relying on a single factor in the prior art, the present invention provides a method and system for monitoring the operating state of a wet washing machine.

[0006] In the first aspect, the present invention provides a method for monitoring the operating state of a wet washing machine, adopting the following technical solution: Obtain multiple characteristic parameters of the wet washing machine and the corresponding state categories of the wet washing machine, and calculate the mutual information entropy between each characteristic parameter and the state category; Use a preset XGboost model to classify the characteristic parameters to obtain the probabilities of each state category of the wet washing machine, calculate the importance coefficient and classification accuracy of each decision tree in the XGboost model, the importance coefficient represents the importance degree of the decision tree, and calculate the weight coefficient of each decision tree, and the weight coefficient is positively correlated with the product of the importance coefficient and the classification accuracy; The feature parameters obtained in real time are classified using a decision tree to obtain a classification result, which is the probability of each state category of the wet washing machine. The classification result is subjected to decision fusion through a voting method to obtain the final decision on the state category of the wet washing machine. Among them, the expression of the final decision is: ; In the formula, is the final decision of the XGboost model on the state category of the wet washing machine, is the total number of decision trees, is the weight coefficient of the k-th decision tree, is the probability that the wet washing machine belongs to the n-th state category obtained by the k-th decision tree.

[0007] Its effect is that by comprehensively analyzing the working state of the wet washing machine using multiple feature parameters, the dependence of the analysis result on a single factor is reduced, and the accuracy and reliability of the analysis result are increased. Moreover, by using a decision tree and performing weighted summation on the classification results of the decision tree to obtain the final decision, the accuracy of the analysis result is further improved.

[0008] Preferably, the expression of the importance coefficient of the decision tree is:

[0009] In the formula, is the importance coefficient of the k-th decision tree; is the total number of feature parameters assigned to the k-th decision tree; is the information entropy of the X-th feature parameter; is the information entropy of the Y-th state category of the wet washing machine; is the mutual information entropy between the X-th feature parameter and the Y-th state category of the wet washing machine.

[0010] Its effect is that by calculating the accuracy of the decision tree, it is convenient to calculate the weight of the decision tree and improve the accuracy of the final decision.

[0011] Preferably, the expression of the mutual information entropy between the feature parameter and the state category of the wet washing machine is:

[0012] In the formula, is the mutual information entropy between the X-th feature parameter and the Y-th state category of the wet washing machine; is the information entropy of the X-th feature parameter; is the information entropy of the Y-th state category of the wet washing machine; H(X, Y) is the joint distribution probability of the X-th feature parameter and the Y-th state category of the wet washing machine.

[0013] Preferably, the expressions of the information entropy of the feature parameter and the information entropy of the state category of the wet washing machine are:

[0014]

[0015] In the formula, is the information entropy of the X-th characteristic parameter; is the characteristic parameter 's marginal distribution probability; is the value set of the characteristic parameter; is the information entropy of the Y-th state category of the wet washing machine; is the category set of the working state of the wet washing machine; is the marginal distribution probability of the Y-th state category of the wet washing machine.

[0016] Preferably, the method for obtaining multiple characteristic parameters of the wet washing machine is as follows: Obtain the sound signal, vibration signal and drum rotation speed signal of the wet washing machine; Perform Fourier transform on the sound signal to obtain the first amplitude, first frequency spectrum and first frequency; Perform Fourier transform on the vibration signal to obtain the second amplitude, second frequency spectrum and second frequency; Perform Fourier transform on the drum rotation speed signal to obtain the third amplitude, third frequency spectrum and third frequency; Use the obtained first amplitude, first frequency spectrum, first frequency, second amplitude, second frequency spectrum, second frequency, third amplitude, third frequency spectrum and third frequency as characteristic parameters.

[0017] The effect is that by converting the sound signal, vibration signal and drum rotation speed signal during the operation of the wet washing machine, multiple characteristic parameters are obtained, which is convenient for analyzing the multiple characteristic parameters, improves the accuracy of the final decision, and is convenient for judging the working state of the wet washing machine.

[0018] Preferably, before performing Fourier transform on the sound signal, vibration signal and drum rotation speed signal, it also includes the step of denoising the sound signal, vibration signal and drum rotation speed signal.

[0019] The effect is that by removing the noise of the signal, the accuracy of the obtained characteristic parameters is further improved.

[0020] Preferably, the method for calculating the classification accuracy of each decision tree in the XGboost model is as follows: Construct a training set and a test set. Both the training set and the test set include multiple characteristic parameters. Set corresponding labels for each characteristic parameter. The label is the state category of the corresponding wet washing machine. Use the training set to train a pre-constructed classification model to obtain the XGboost model; Input the characteristic parameters in the test set into the XGboost model. For each decision tree, obtain the status category of the wet washing machine, count the proportion of the number of status categories that are the same as the actual status category, and use the proportion as the classification accuracy of the corresponding decision tree.

[0021] Preferably, the weight coefficient expression of each decision tree is:

[0022] In the formula, is the weight coefficient of the k-th decision tree, is the accuracy of the classification of the working status of the wet washing machine obtained by the k-th decision tree, is the importance coefficient of the k-th decision tree; K is the total number of decision trees in the model.

[0023] Its effect is that by calculating the weight coefficient of each decision tree, it is convenient to understand the importance degree of the decision tree and improve the accuracy of the final decision.

[0024] In the second aspect, the present invention provides a wet washing machine operation status monitoring system, adopting the following technical solution: A wet washing machine operation status monitoring system includes: a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned wet washing machine operation status monitoring method is implemented.

[0025] The present invention has the following technical effects: By comprehensively analyzing the working status of the wet washing machine using multiple characteristic parameters, compared with the results of single-factor analysis, the dependence of the analysis results on a single factor is reduced, the accuracy and reliability of the analysis results are increased, and by using decision trees, according to the weight coefficients of the decision trees, the analysis and classification results of the decision trees are weighted and summed to obtain the final decision, further improving the accuracy of the analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] By referring to the accompanying drawings and reading the following detailed description, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become easily understood. In the drawings, several embodiments of the present invention are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0027] Figure 1 is the flowchart of the method for monitoring the operation status of a wet washing machine according to an embodiment of the present invention.

[0028] Figure 2 is the structural block diagram of a wet washing machine operation status monitoring system according to the present invention. DETAILED DESCRIPTION

[0029] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0030] It should be understood that when terms such as "first" and "second" are used in the claims, the description, and the drawings of the present invention, they are only used to distinguish different objects and not to describe a specific order. The terms "including" and "comprising" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0031] An embodiment of the present invention discloses a method for monitoring the operating state of a wet washing machine. Refer to Figure 1 , and the following steps are included, specifically as follows: S1: Obtain multiple characteristic parameters of the wet washing machine and the corresponding state categories of the wet washing machine.

[0032] Obtain the working indicators and state categories of the wet washing machine at the same moment. The working indicators include sound signals, vibration signals, and drum rotation speed signals. Denoise the sound signals, vibration signals, and drum rotation speed signals to obtain the denoised sound signals, vibration signals, and drum rotation speed signals. Then, perform Fourier transform on the denoised sound signals to obtain the first amplitude, the first frequency spectrum, and the first frequency; perform Fourier transform on the denoised vibration signals to obtain the second amplitude, the second frequency spectrum, and the second frequency; perform Fourier transform on the denoised drum rotation speed signals to obtain the third amplitude, the third frequency spectrum, and the third frequency; use the obtained first amplitude, the first frequency spectrum, the first frequency, the second amplitude, the second frequency spectrum, the second frequency, the third amplitude, the third frequency spectrum, and the third frequency as characteristic parameters. During the operation of the wet washing machine, the sound signals, vibration signals, and drum rotation speed signals can reflect the working state of the wet washing machine. For example, before the wet washing machine malfunctions, abnormal sounds will be emitted, irregular vibrations will also occur, and the drum rotation speed will suddenly become faster or slower. Therefore, the working state of the wet washing machine can be understood in real time through these signals.

[0033] Obtain the state categories of the wet washing machine at the same moment. The state categories include normal, to be maintained, and faulty; it can be understood that each characteristic parameter corresponds to a state category. Among them, normal indicates that the wet washing machine has good working performance and can be in a normal operating state; to be maintained indicates that the working performance of the wet washing machine is average and maintenance is required to improve the working performance of the wet washing machine; faulty indicates that the wet washing machine is damaged and cannot work normally.

[0034] S2: Calculate the mutual information entropy between each feature parameter and the state category.

[0035] S21: Obtain the feature parameters and state categories at multiple moments to construct a data set. Taking the data in the data set as an example, calculate the joint frequency of the feature parameter and the state category , and calculate the marginal frequency of the feature parameter respectively and the marginal frequency of the state category . Then, calculate the marginal distribution probability of the feature parameter, the marginal distribution probability of the state category, and the joint distribution probability of the feature parameter and the working state of the wet washing machine. The expressions are: , , ; In the formula, is the total number of feature parameters in the data set; is the marginal distribution probability of the Xth feature parameter; is the marginal distribution probability of the Yth state category of the wet washing machine; is the joint distribution probability of the Xth feature parameter and the Yth state category of the wet washing machine.

[0036] S22: Calculate the information entropy of the feature parameter, the information entropy of the state category, and the joint distribution entropy of the feature parameter and the state category of the wet washing machine.

[0037] The expressions are: ; ; ; In the formula, is the information entropy of each Xth feature parameter; is the information entropy of the Yth state category of the wet washing machine; is the joint distribution entropy of the Xth feature parameter and the Yth state category of the wet washing machine, is the set of wet washing machine state categories; is the set of feature parameters; is the marginal distribution probability of the Xth feature parameter; is the marginal distribution probability of the Yth state category of the wet washing machine; is the joint distribution probability of the Xth feature parameter and the Yth state category of the wet washing machine.

[0038] The expression for the mutual information entropy between the feature parameter and the state category is:

[0039] I(X;Y) is the mutual information entropy between the X-th characteristic parameter and the Y-th state category of the wet washing machine. The higher the mutual information entropy value, the more important the characteristic parameter is for distinguishing different working states.

[0040] S3: Use the preset XGboost model to classify the characteristic parameters to obtain the probabilities of each state category of the wet washing machine, and calculate the importance coefficient and classification accuracy of each decision tree in the XGboost model. The importance coefficient represents the importance degree of the decision tree.

[0041] Construct a classification model and construct a training set and a test set. Both the training set and the test set include multiple characteristic parameter sequences. Each characteristic parameter sequence contains multiple characteristic parameters. Set corresponding labels for each characteristic parameter sequence. The label is the state category of the corresponding wet washing machine. It can also be understood that the characteristic parameters within the characteristic parameter sequence belong to the same state category. Use the training set to train the pre-constructed classification model to obtain the XGboost model; input the characteristic parameters in the test set into the XGboost model. For each decision tree, obtain the state category of the wet washing machine, and count the proportion of the number of state categories that are the same as the actual state category. Take the proportion as the classification accuracy of the corresponding decision tree.

[0042] In the process of calculating the classification accuracy of the decision tree, take the state category with the highest output probability as the final result. Exemplarily, for one of the decision trees, the classification result of the data is (0.6, 0.3, 0.1), indicating that the probability of the normal working state of the wet washing machine is 0.6, the probability of maintenance required is 0.3, and the probability of failure is 0.1. Then record the state category of the wet washing machine obtained by this classification as normal. Suppose in 10 test processes of the decision tree, 8 classification results are the same as the label, then the accuracy rate of the corresponding decision tree is 0.8.

[0043] Normalize the information entropy of the characteristic parameters, the information entropy of the state categories, and the mutual information entropy, and use the normalized results to calculate the importance coefficient of the decision tree. The expression of the importance coefficient of the decision tree is: ; In the formula, is the importance coefficient of the k-th decision tree; is the total number of characteristic parameters assigned to the k-th decision tree; is the information entropy of the X-th characteristic parameter; is the information entropy of the Y-th state category of the wet washing machine; is the mutual information entropy between the X-th characteristic parameter and the Y-th state category of the wet washing machine. The larger the importance coefficient, the more important the corresponding decision tree is in the classification process.

[0044] S4: Calculate the weight coefficient of each decision tree. The weight coefficient is positively correlated with the product of the importance coefficient and the classification accuracy.

[0045] The weight coefficient expression of each decision tree is as follows:

[0046] In the formula, is the weight coefficient of the k-th decision tree, is the accuracy of the state category of the washing machine obtained by the k-th classification decision, is the importance coefficient of the k-th decision tree, and K is the total number of decision trees in the model. The larger the weight coefficient of the decision tree, the more important the result output by the decision tree in the classification process.

[0047] S5: Use the decision tree to classify the real-time obtained parameter set to obtain a classification result. The classification result is the probability of each state category of the washing machine. The voting method is used to perform decision fusion on the classification result to obtain the final decision of the state category of the washing machine.

[0048] The expression of the final decision is: ; In the formula, is the final decision of the XGboost model on the state category of the washing machine, K is the total number of decision trees, is the weight coefficient of the k-th decision tree, is the probability that the washing machine belongs to the n-th state category obtained by the k-th decision tree.

[0049] Exemplarily, in this embodiment, the state categories are divided into: 1 class normal, 2 class to be maintained, 3 class faulty, where represents the cumulative sum of the weight coefficients of all decision trees and the probability of the normal state category, represents the cumulative sum of the weight coefficients of all decision trees and the probability of the state category to be maintained, represents the cumulative sum of the weight coefficients of all decision trees and the probability of the faulty state category. For example, the weight coefficient of the first tree is 0.6, and the output result is (0.6, 0.3, 0.1), the weight coefficient of the second tree is 0.3, and the output result is (0.7, 0.2, 0.1), the weight coefficient of the third tree is 0.1, and the output result is (0.3, 0.5, 0.2), then, The result of is: 0.6×0.6 + 0.3×0.7 + 0.1×0.3 = 0.60; The result of is 0.6×0.3 + 0.3×0.2 + 0.1×0.5 = 0.29; The result is 0.6×0.1 + 0.3×0.1 + 0.1×0.2 = 0.11; it indicates that the probability of the decision result being normal is 0.6, the probability of being to be maintained is 0.29, and the probability of failure is 0.11. Since the probability value of being normal is the largest, the final decision result is normal, indicating that the working state of the wet washing machine is likely to be normal during the current time period.

[0050] Suppose the final classification result is (0.2, 0.1, 0.7), which means that the wet washing machine is likely to fail during the current time period, and the operation should be stopped to detect the equipment and eliminate the fault.

[0051] Combined with Figure 2 As shown, the embodiment of the present invention also discloses a wet washing machine operation state monitoring system, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a wet washing machine operation state monitoring method according to the present invention is implemented.

[0052] The above system also includes other components well-known to those skilled in the art such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be described in detail here.

[0053] In the present invention, the aforementioned memory can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component. For example, a computer-readable storage medium can be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory (RRAM), a dynamic random access memory (DRAM), a static random access memory (SRAM), an enhanced dynamic random access memory (EDRAM), a high bandwidth memory (HBM), a hybrid memory cube (HMC), etc., or any other medium that can be used to store the required information and can be accessed by an application program, module, or both. Any such computer storage medium can be part of the device or accessible or connectable to the device.

[0054] Although this specification has shown and described multiple embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will think of many changes, alterations, and alternative ways without departing from the spirit and concept of the present invention. It should be understood that various alternatives to the embodiments of the present invention described herein may be employed in the practice of the present invention.

[0055] The above are all preferred embodiments of the present invention, and the protection scope of the present invention is not limited thereby. Therefore, all equivalent changes made according to the structure, shape, and principle of the present invention shall be covered within the protection scope of the present invention.

Claims

1. A method for monitoring the operating status of a wet cleaning machine, characterized in that: Includes steps: Acquire multiple characteristic parameters of the wet cleaning machine and the corresponding state category of the wet cleaning machine, and calculate the mutual information entropy between each characteristic parameter and the state category; The preset XGboost model is used to classify the feature parameters to obtain the probability of each state category of the wet cleaning machine. The importance coefficient and classification accuracy of each decision tree in the XGboost model are calculated. The importance coefficient represents the importance of the decision tree. The weight coefficient of each decision tree is calculated. The weight coefficient is positively correlated with the product of the importance coefficient and the classification accuracy. The decision tree is used to classify the feature parameters obtained in real time to obtain the classification results. The classification results are the probabilities of each state category of the wet cleaning machine. The classification results are fused by the voting method to obtain the final decision of the wet cleaning machine state category. The final decision expression is: ; In the formula, is the final decision of the XGboost model on the wet cleaning machine status category, is the total number of decision trees, is the weight coefficient of the kth decision tree, The probability that the wet cleaning machine belongs to the nth state category obtained by the kth decision tree.

2. A method for monitoring the operating status of a wet cleaning machine according to claim 1, characterized in that: The importance coefficient expression of the decision tree is: In the formula, is the importance coefficient of the kth decision tree; is the total number of feature parameters assigned to the kth decision tree; is the information entropy of the Xth characteristic parameter; is the information entropy of the Yth state category of the wet cleaning machine; is the mutual information entropy between the Xth characteristic parameter and the Yth state category of the wet cleaning machine.

3. A method for monitoring the operating status of a wet cleaning machine according to claim 1, characterized in that: The expression of mutual information entropy between characteristic parameters and state category of wet cleaning machine is: In the formula, is the mutual information entropy between the Xth characteristic parameter and the Yth state category of the wet cleaning machine; is the information entropy of the Xth characteristic parameter; is the information entropy of the Yth state category of the wet cleaning machine; H(X,Y) is the joint distribution probability of the Xth characteristic parameter and the Yth state category of the wet cleaning machine.

4. A method for monitoring the operating status of a wet cleaning machine according to claim 3, characterized in that: The information entropy of characteristic parameters and the information entropy of the state category of the wet cleaning machine are expressed as: In the formula, is the information entropy of the Xth characteristic parameter; is the marginal distribution probability of the Xth feature parameter; is the value set of feature parameters; is the information entropy of the Yth state category of the wet cleaning machine; A category collection of the working status of the wet cleaning machine; is the marginal distribution probability of the Yth state category of the wet cleaning machine.

5. A method for monitoring the operating status of a wet cleaning machine according to claim 1, characterized in that: The method for obtaining multiple characteristic parameters of the wet cleaning machine is: Acquire the sound signal, vibration signal and drum speed signal of the wet cleaning machine; Performing Fourier transformation on the sound signal to obtain a first amplitude, a first spectrum and a first frequency; Performing Fourier transformation on the vibration signal to obtain a second amplitude, a second spectrum and a second frequency; Performing Fourier transformation on the drum speed signal to obtain a third amplitude, a third spectrum and a third frequency; The obtained first amplitude, first spectrum, first frequency, second amplitude, second spectrum, second frequency, third amplitude, third spectrum and third frequency are taken as characteristic parameters.

6. A method for monitoring the operating status of a wet cleaning machine according to claim 5, characterized in that: Before Fourier transforming the sound signal, the vibration signal and the drum speed signal, the method further includes denoising the sound signal, the vibration signal and the drum speed signal.

7. A method for monitoring the operating status of a wet cleaning machine according to claim 1, characterized in that: The method to calculate the classification accuracy of each decision tree in the XGboost model is: Constructing a training set and a test set, each of which includes multiple feature parameter sequences, each parameter sequence includes multiple feature parameters, setting a corresponding label for each feature parameter sequence, the label is the state category of the corresponding wet cleaning machine, and using the training set to train the pre-constructed classification model to obtain an XGboost model; The feature parameters in the test set are input into the XGboost model. For each decision tree, the state category of the wet cleaning machine is obtained, and the number of state categories that are the same as the actual state category is counted, and the number ratio is used as the classification accuracy of the corresponding decision tree.

8. A method for monitoring the operating status of a wet cleaning machine according to claim 1, characterized in that: The weight coefficient expression of each decision tree is: In the formula, is the weight coefficient of the kth decision tree, is the accuracy of the wet cleaning machine status category obtained by the k-th decision tree, is the importance coefficient of the kth decision tree; K is the total number of decision trees in the model.

9. A wet cleaning machine operation status monitoring system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for monitoring the operating status of a wet cleaning machine according to any one of claims 1 to 8 is implemented.

Images