TWI900902B - Method and electronic deivce for evaluating oil degradation - Google Patents

Abstract

A method is provided for evaluating oil degradation, which includes: obtaining multiple test values related to the oil; categorizing these test values into multiple classes including a chemical class, a wear class, and a contamination class; for each test value in each class, executing a clustering algorithm to calculate a corresponding cluster code; for each class, calculating a class indicator based on the corresponding cluster codes of the test values; obtaining the class indicator that represents the most severe level of degradation among all classes as a comprehensive oil indicator; and determining whether to replace the oil based on the comprehensive oil indicator.

Description

Oil degradation assessment method and electronic device

This disclosure relates to a method and electronic device that can objectively and automatically assess the degree of oil degradation.

Lubricants have many important functions in equipment, a few of which are listed below. Reducing friction and wear: One of the main functions of lubricants is to form a lubricating film between two contacting surfaces to reduce friction and wear. Heat transfer: Lubricants can transfer heat from heat-generating areas (such as the engine cylinder) to the cooling system or the outside world, thereby playing a cooling role. Anti-corrosion and corrosion resistance: Many lubricants contain special additives that can prevent metal parts from rusting or corrosion due to moisture or other corrosive substances. Sealing: Lubricants can fill certain tiny gaps, acting as seals to prevent the leakage of gas or liquid. Cleanliness: Lubricating oil can remove dirt, debris, etc. from the engine or other mechanical parts to keep the machine clean. Transmitting power: In some hydraulic systems or transmissions, lubricating oil not only provides lubrication, but is also used to transmit power. Damping vibrations: Due to its viscosity, lubricating oil can also absorb or reduce vibrations in mechanical movement to a certain extent. Improving efficiency: Due to the reduction in friction, mechanical equipment operates more smoothly, thereby improving the efficiency of energy conversion. In conventional technology, after obtaining a lot of data through oil analysis, it is manually determined whether the lubricating oil should be replaced. However, this approach relies on human experience and is not objective and cannot be automated.

The disclosed embodiments provide a method for evaluating oil degradation, performed by an electronic device. This method includes obtaining multiple test values related to the oil; classifying the test values into multiple categories, including chemical, wear, and contamination categories; executing a clustering algorithm to calculate a corresponding cluster code for each test value in each category; calculating a category index for each category based on the cluster code of the corresponding test value; obtaining the category index representing the most severe degradation among all categories as a comprehensive oil index; and determining whether to replace the oil based on the comprehensive oil index.

In some embodiments, the clustering algorithm is a balanced iterative reduction clustering method. When the test value for clustering is acid value or cleanliness, the clustering code is one of three codes. When the test value for clustering is not acid value or cleanliness, the clustering code is one of five codes.

In some embodiments, the step of calculating the category index includes: multiplying the value corresponding to each clustering code by a weight to obtain a product, and then adding the products corresponding to a category to obtain the category index.

In some embodiments, the oil degradation assessment method further includes: obtaining multiple training data, each training data including an inspection value and an abnormality indicator, wherein the abnormality indicator indicates whether the corresponding equipment is abnormal within a preset time; calculating a training category indicator for each training data; and obtaining a probability distribution of the corresponding category based on the training category indicator and the abnormality indicator, thereby calculating the failure probability.

In some embodiments, the oil degradation assessment method further includes: providing a user interface, wherein the user interface includes a radar distribution diagram or a curve diagram, wherein the radar distribution diagram includes category indicators, and the curve diagram includes category indicators that change over time.

From another perspective, embodiments of the present disclosure provide an electronic device comprising a memory and a processor. The memory stores a plurality of instructions. The processor is communicatively connected to the memory to execute the instructions to implement the aforementioned oil degradation assessment method.

In the above-mentioned device and method, the degree of oil degradation can be objectively and accurately evaluated.

In order to make the above features and advantages of the present invention more clearly understood, embodiments are given below and described in detail with reference to the accompanying drawings.

The terms “first,” “second,” etc. used herein do not particularly refer to order or sequence, but are only used to distinguish elements or operations described with the same technical terms.

Figure 1 is a schematic diagram of an electronic device according to one embodiment. Referring to Figure 1 , electronic device 100 may be a smartphone, tablet computer, personal computer, laptop computer, server, distributed computing machine, cloud server, industrial computer, or any other electronic device with computing capabilities, although the present invention is not limited thereto. Electronic device 100 includes a processor 110 and a memory 120 . Processor 110 is communicatively connected to memory 120 . This communication connection may be achieved via any wired or wireless communication means, or may be achieved via the Internet. The processor 110 can be a central processing unit (CPU), a microprocessor, a microcontroller, an application-specific integrated circuit (ASIC), or the like. The memory 120 can be a random access memory (RAM), a read-only memory (ROM), a flash memory, a floppy disk, a hard disk, an optical disk, a flash drive, a magnetic tape, or a database accessible via the internet. Multiple instructions are stored therein, and the processor 110 executes these instructions to complete an oil degradation assessment method, which is described in detail below.

Figure 2 is a flow chart illustrating a method for evaluating oil degradation according to one embodiment. Referring to Figure 2 , in step 201, multiple test values for the oil are obtained. The oil may be refrigeration oil, hydraulic oil, gasoline oil, gear oil, etc. Test values may include, for example, viscosity, water content, acid value, cleanliness, various metal contents, additive content, etc., but this disclosure is not limited thereto. In some embodiments, the additive content includes sub-items such as zinc content, calcium content, and magnesium content, and the metal content also includes sub-items such as iron content, copper content, lead content, and tin content.

In step 202, the test values are categorized into multiple categories, including chemical properties, wear properties, and contamination. For example, chemical properties include viscosity and acid value; wear properties include metal content and clarity; and contamination properties include additive content and water content. In this embodiment, each category of the oil is tested for abnormalities.

In step 203, for each test value in each category, a clustering algorithm is executed to calculate the corresponding clustering code. The number of clusters can be set arbitrarily based on experimental results. For example, when the test value to be clustered is acid value or cleanliness, there are three clusters, and the clustering codes of these three clusters are "excellent", "acceptable", and "degraded". When the test value to be clustered is not acid value or cleanliness, there are five clusters, and the clustering codes of these five clusters are "excellent", "good", "acceptable", "poor", and "degraded". Here, text is used as the clustering code, but in other embodiments, any number, text, or binary character can also be used as the clustering code, and the present disclosure is not limited to this.

In some embodiments, the clustering algorithm is Balanced Iterative Reducing and Clustering Using Hierarchies (BIRCH). However, in other embodiments, non-supervised clustering algorithms such as the k-means clustering algorithm and DBSCAN (Density-based spatial clustering of applications with noise) may also be used, and the present disclosure is not limited thereto. During the training phase, each training data set contains all the test values of the oil product, and each test value is manually assigned a label (i.e., the clustering code mentioned above). During training, the parameters in the model are adjusted to minimize the set error. This error can be mean square error (MSE), mean absolute error (MAE), cross-entropy, Huber loss function, Log-Cosh loss function, etc., but this disclosure is not limited to this.

Next, in step 204, a category index is calculated for each category based on the clustering code corresponding to the test value. Each clustering code is assigned a value and a weight. The value corresponding to each clustering code is multiplied by the corresponding weight to obtain a product. All corresponding products within the same category are then summed to obtain the corresponding category index. For example, in the chemical property category, acid value is divided into three clustering codes: "Excellent," "Fair," and "Degraded." The corresponding three values are 1, 2, and 3, respectively. In other words, a larger value indicates a greater degree of degradation, and the corresponding three weights are 1, 1, and 1.5. For the clustering code "Degraded," a higher weight must be assigned to emphasize the degree of degradation. In addition, the five viscosity cluster codes also have corresponding numerical values and weights. For example, the numerical values corresponding to the five cluster codes "Excellent," "Good," "Acceptable," "Poor," and "Degraded" are 1, 2, 3, 4, and 5, respectively. Similarly, larger numerical values indicate greater degradation, and the corresponding weights are 1, 1, 1, 1.25, and 1.5. For the cluster codes "Poor" and "Degraded," higher weights are assigned to emphasize the degree of degradation. Next, the weight sum is calculated: Chemical Class Index = Numerical Value Corresponding to Viscosity Cluster Code x Weight Corresponding to Viscosity Cluster Code + Numerical Value Corresponding to Acid Value Cluster Code x Weight Corresponding to Acid Value Cluster Code. For example, if an oil is classified as "acceptable" for acid value and "degraded" for viscosity, the chemical classification index is 2x1+5x1.5=9.5.

In the above embodiment, the category index is a continuous numerical value, but in some embodiments, the continuous numerical value can also be converted into a discrete category. For example, if the calculated category index is greater than or equal to 0 and less than 2, it can be converted to the "Excellent A"category; if the calculated category index is greater than or equal to 2 and less than 4, it can be converted to the "Good B" category, and so on. For each inspection item in the wearability category and the contamination category, a corresponding numerical value and weight can be set for the grouping code. Similarly, the category index of the wearability category (which can be continuous or discrete) and the category index of the contamination category (which can be continuous or discrete) can be calculated by summing the weights. The above values and weights can be set through human experience, or can also be set through experiments and statistical methods. This disclosure is not limited to these values. In this embodiment, the continuous values are converted into discrete category indicators through the following Table 1. Project Excellent A Good B Fair C Poor D DeteriorationE Chemical Class 0~2 2~4 4~6 6~7 7 or more Pollution level categories 0~6 7~9 10~13 14~17 17 and above Abrasiveness category 0~7 8~10 11~15 16~19 19 and above Damage probability 0 50% 60% 80% 90% Table 1

The last column of Table 1 also shows the probability of failure. Failure is defined as the occurrence of a failure in a device due to an oil abnormality or an oil abnormality within a preset period of time (e.g., three months) after the oil test. In each training data set, in addition to the oil test value, there is also an abnormality indicator, which indicates whether the corresponding device has experienced an abnormality within the preset time period. After all training data is collected, the aforementioned calculation method can be used to calculate the category index (also called the training category index) for each category in each training data set. Next, based on these training category indexes and the abnormality index, the probability distribution of the corresponding category can be obtained, thereby calculating the probability of failure. For example, a probability density function (PDF) can be obtained, expressed as P(x), where x is the class index and P() is the probability of damage. Each class has its own PDF. Once the class index is calculated, it is substituted into the PDF and the probability of damage can be calculated using integration or other methods. In some embodiments, the class index for each class can also be expressed as a probability of damage.

Next, proceed to step 205 to obtain the category index representing the most severe degradation as the comprehensive oil quality index. In this embodiment, there are three category indexes, corresponding to the chemical category, the wear category, and the contamination category. Since any one category indicates that the oil has degraded, it may be unsuitable for continued use. Therefore, the category index representing the most severe degradation is used as the comprehensive oil quality index. If a continuous numerical value is used as the category index, for example, in the above embodiment, a larger numerical value indicates a more severe degradation, then the category index with the largest numerical value among all categories can be used as the comprehensive oil quality index. If discrete categories are used as category indexes, category indexes such as "Degraded E" and "Poor D" are preferred. For example, if the category index of the chemical category is "Deteriorated E", the category index of the wear category is "Acceptable C", and the category index of the abrasion category is "Good B", then "Deteriorated E" is used as the comprehensive oil index.

Next, step 206 determines whether the oil should be replaced based on the oil quality index. If the oil quality index is a continuous value, a threshold value can be set. If the oil quality index exceeds this threshold value, the oil is determined to be replaced; otherwise, no replacement is required. If the oil quality index is a discrete category, the oil quality index is determined to be a preset code (for example, whether it is "degraded"). If so, the oil is determined to be replaced; otherwise, no replacement is required. After determining that the oil needs to be replaced, a message can be sent to relevant personnel. This message can be displayed on a user interface and can be presented in the form of text, numbers, pictures, colors, or sounds. Alternatively, the message can be sent to a mobile phone used by the relevant personnel, but this disclosure is not limited thereto.

In some embodiments, electronic device 100 can also display a radar distribution map or graph on the user interface to represent various category indicators. Figure 3 is a schematic diagram illustrating a radar map according to one embodiment. Referring to Figure 3 , the three corners of radar map 300 represent the three categories of contamination, wear, and chemistry. The category indicators are presented discretely, with different colors marking "Excellent," "Good," "Acceptable," "Poor," and "Degraded." This allows operators to quickly identify degradation in the chemistry category indicator from radar map 300.

Figure 4 is a schematic diagram illustrating a graph according to one embodiment. Class indicators are presented using continuous numerical values. The horizontal axis of graph 400 represents time, and the vertical axis represents the class indicator. Therefore, graph 400 includes class indicators that change over time. Curve 401 represents the chemical class, curve 402 represents the wear class, and curve 403 represents the contamination class. Furthermore, line 404 represents a critical value. Exceeding this critical value indicates the need for oil replacement. In this example, curve 401, representing the chemical class, exceeds the critical value within warning zone 410, indicating that the oil has deteriorated, with chemical properties being the primary cause of the degradation.

The above-mentioned comprehensive oil quality indicators can be used to determine when oil needs to be replaced and which factor (chemical, wear, or contamination) is causing oil degradation. In some embodiments, oil can be replaced earlier when oil degradation is imminent. Conventional methods use manual methods to determine whether to replace oil. This can result in wasteful and environmentally unfriendly changes due to premature replacements, or equipment malfunctions due to late replacements. The disclosed method can accurately determine when oil replacement is necessary.

Although the present invention has been disclosed above by way of embodiments, they are not intended to limit the present invention. Any person having ordinary skill in the art may make slight modifications and improvements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the scope of the attached patent application.

100: Electronic device 110: Processor 120: Memory 201-206: Steps 300: Radar graph 400: Curve graph 401-403: Curves 404: Line 410: Warning area

Figure 1 is a schematic diagram illustrating an electronic device according to one embodiment. Figure 2 is a flow chart illustrating a method for evaluating oil degradation according to one embodiment. Figure 3 is a schematic diagram illustrating a radar image according to one embodiment. Figure 4 is a schematic diagram illustrating a curve graph according to one embodiment.

201~206: Steps

Claims (10)

A method for assessing oil degradation, performed by an electronic device, comprises: obtaining multiple test values for an oil product; classifying the test values into multiple categories, the categories comprising a chemical property category, a wear property category, and a contamination degree category; executing a clustering algorithm for each of the test values within each of the categories to calculate a corresponding cluster code; calculating a category index for each of the categories based on the cluster codes of the corresponding test values; obtaining the category index representing the most severe degradation within the categories as a comprehensive oil product index; and determining whether to replace the oil product based on the comprehensive oil product index. The oil degradation assessment method as described in claim 1, wherein the clustering algorithm is a balanced iterative reduction clustering method; When the test value of the cluster is acid value or purity, the clustering code is one of three codes; When the test value of the cluster is not the acid value or purity, the clustering code is one of five codes. The oil degradation assessment method as described in claim 1, wherein for each of the categories, the step of calculating the category index based on the grouping codes corresponding to the inspection values comprises: Multiplying a value corresponding to each of the grouping codes by a weight to obtain a product, and then summing the products corresponding to one of the categories to obtain the category index. The oil degradation assessment method as described in claim 1 further includes: obtaining a plurality of training data, wherein each of the training data includes the test values and an abnormality indicator, the abnormality indicator indicating whether a corresponding device is abnormal within a predetermined time period; calculating a plurality of training category indicators for each of the training data; and obtaining a probability distribution corresponding to one of the categories based on the training category indicators and the abnormality indicators, thereby calculating a failure probability. The oil degradation assessment method of claim 1 further comprises: Providing a user interface, the user interface comprising a radar distribution graph or a curve graph, the radar distribution graph comprising the category indicators, and the curve graph comprising the category indicators changing over time. An electronic device comprises: A memory storing a plurality of instructions; and A processor communicatively connected to the memory for executing the instructions to complete a plurality of steps: Obtaining a plurality of test values related to an oil product; Classifying the test values into a plurality of categories, the categories comprising a chemical category, a wear category, and a contamination category; Executing a clustering algorithm for each of the test values in each of the categories to calculate a corresponding clustering code; For each of the categories, calculating a category index based on the clustering codes of the corresponding test values; Obtaining the category index representing the most severe deterioration among the categories as a comprehensive oil product index; and Determine whether to replace the oil based on its comprehensive indicators. The electronic device of claim 6, wherein the clustering algorithm is a balanced iterative reduction clustering method, wherein, when the test value of the cluster is acid value or cleanliness, the clustering code is one of three codes, wherein, when the test value of the cluster is not the acid value or the cleanliness, the clustering code is one of five codes. The electronic device of claim 6, wherein for each of the categories, the step of calculating the category index based on the grouping codes corresponding to the test values comprises: Multiplying each of the grouping codes by a weight to obtain a product, and then summing the products corresponding to one of the categories to obtain the category index. The electronic device of claim 6, wherein the steps further include: obtaining a plurality of training data, wherein each of the training data includes the test values and an abnormality indicator, the abnormality indicator indicating whether a corresponding device is abnormal within a predetermined time period; calculating a plurality of training category indicators for each of the training data; and obtaining a probability distribution corresponding to one of the categories based on the training category indicators and the abnormality indicators, thereby calculating a failure probability. The electronic device of claim 6, wherein the steps further include: Providing a user interface, the user interface comprising a radar distribution graph or a curve graph, the radar distribution graph comprising the category indicators, and the curve graph comprising the category indicators changing over time.