Refrigerator service life prediction method and device, computer equipment and storage medium
By building a prediction model based on historical equipment data, the lifespan of the equipment to be tested is quickly predicted, and the problem of life prediction in the existing technology is solved, and efficient and accurate life prediction is achieved, which is conducive to the rapid update and iteration of products.
Patent Information
- Application Number
- CN202411578708.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-07
- Publication Date
- 2025-02-25
AI Technical Summary
The existing Stirling refrigerator life prediction method takes a long time and cannot reflect the life of batch machines in a timely manner, which is not conducive to product updates and iterations.
By obtaining the historical data sets of multiple historical devices for preprocessing, constructing historical feature data, and finding target devices with similarity to the equipment to be tested is greater than a similar threshold, so as to build a prediction model based on the experimental data of the target device and predict the life of the equipment to be tested.
It realizes rapid and accurate life prediction of the equipment to be tested without long life experiments, reduces the testing cost, has good timeliness, and is conducive to product updates and iterations.
Smart Images
Figure CN119514353A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of equipment control technology, and in particular to a refrigerator life prediction method, device, computer equipment, and storage medium. Background Art
[0002] The existing life prediction method of Stirling refrigerator mainly judges the life of Stirling refrigerator by sampling batches of machines for accelerated life verification. However, life experiments usually take a long time and cannot reflect the life of batch machines in a timely manner, which is not conducive to product updates and iterations. Summary of the invention
[0003] Based on this, it is necessary to provide a refrigerator life prediction method, device, computer equipment, computer readable storage medium and computer program product that can quickly predict the life of batch refrigerators in response to the above technical problems.
[0004] In a first aspect, the present application provides a refrigerator life prediction method, comprising: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0005] In one embodiment, the preprocessing of the historical data set to obtain an optimized data set includes: Performing label conversion on the lifespan data in the historical data set to obtain a converted data set; Anomaly detection and missing fill are performed on the converted data set to obtain the optimized data set.
[0006] In one embodiment, performing label conversion on the lifespan data in the historical data set to obtain a converted data set includes: Find the range interval to which the life span data belongs; The life data is mapped to the label value corresponding to the range interval to obtain the converted data set. In one embodiment, constructing historical feature data based on the optimized data set includes: Screening the optimized data set based on preset indicators to obtain screening results, wherein the preset indicators include current data indicators; The historical feature data is constructed based on the screening result.
[0007] In one embodiment, the screening of the optimized data set based on the preset index to obtain the screening result includes: Obtaining a first-order difference value of a preset index of each element in the optimized data set; If the first-order difference value of the current element is greater than the valid threshold, the current element is determined as a valid element; A set of all the valid elements is determined as the screening result.
[0008] In one embodiment, based on the historical feature data, obtaining a target device having a similarity with the device to be tested greater than a similarity threshold from a historical database includes: Based on the clustering algorithm and the historical feature data, a target device whose similarity with the device to be tested is greater than the similarity threshold is searched from the historical database.
[0009] In a second aspect, the present application also provides a refrigerator life prediction device, comprising: A preprocessing module, used to obtain historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; A construction module, used for constructing historical feature data based on the optimized data set; An acquisition module, configured to acquire, from a historical database, a target device whose similarity to the device to be tested is greater than a similarity threshold based on the historical feature data; A prediction module is used to build a prediction model based on the experimental data of the target device, and predict the life of the device under test based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0010] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0011] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0012] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0013] The above refrigerator life prediction method, device, computer equipment, computer readable storage medium and computer program product find the historical equipment most similar to the equipment to be tested, and then build a prediction model based on the historical experimental data of the historical equipment, and predict the life of the equipment to be tested through the prediction model. It can quickly predict the life of the equipment to be tested by machine learning, and can obtain the expected life of the equipment to be tested without long life experiments. It has good timeliness and reduces testing costs, which is conducive to product updates and iterations. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0015] Figure 1 A schematic diagram of a flow chart of a refrigerator life prediction method in one embodiment; Figure 2 is a structural block diagram of a refrigerator life prediction device in one embodiment; Figure 3 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION
[0016] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0017] The integrated rotary Stirling refrigerator mainly provides an effective cooling source for the low-temperature detector. Its life span will directly affect the service life of the infrared remote sensing equipment. Therefore, the long life of the Stirling refrigerator is the key technology for its application. The existing Stirling refrigerator performance detection system mainly judges the life of the Stirling refrigerator by performing accelerated life verification on batch machines. Although this method can estimate the life of batch refrigerators, in actual production, it may take nearly 10 months of detection cycle. This cycle time is too long and cannot reflect the life of batch machines in a timely manner, which is not conducive to product updates and iterations.
[0018] Based on this, this application proposes a life prediction method based on machine learning, which can quickly obtain the machine life status of each production batch.
[0019] Specifically, Figure 1 As shown, in an exemplary embodiment, a refrigerator life prediction method is provided, including the following steps 101 to 104. Among them: Step 101, obtaining historical data sets of multiple historical devices, and preprocessing the historical data sets to obtain optimized data sets; The original historical data set includes a variety of data, and there may be data anomalies or missing data. Therefore, the original data needs to be preprocessed to remove and fill in the outliers to obtain an optimized data set for subsequent feature extraction and modeling. In addition, there is a special variable called life data in the original historical data set. In the commonly used system reliability judgment, the concept of Mean Time To Failure (MTTF) is usually used to represent the life of the equipment. However, the MTTF value is usually a large value, which will cause inconvenience in data processing. Therefore, this application also performs special processing on the life data during preprocessing. Please see below for the specific processing method.
[0020] Step 102, constructing historical feature data based on the optimized data set; In order to facilitate subsequent matching and modeling, it is necessary to construct features for the optimized data set. In addition, when constructing features, it is also necessary to screen based on some indicators to ensure the validity of the data.
[0021] Step 103, based on the historical feature data, obtaining a target device whose similarity with the device to be tested is greater than a similarity threshold from a historical database; This step realizes the matching of the device under test and the historical device, finds the target device that is closest to the device under test from the historical device, and then builds a model based on the experimental data of the target device. The prediction model obtained in this way can also predict the life of the device under test more accurately.
[0022] Step 104, constructing a prediction model based on the experimental data of the target device, and predicting the life of the device under test based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0023] The prediction model constructed based on the experimental data of the target device can effectively predict the life of the device under test when the collected data of the device under test is input.
[0024] In the above refrigerator life prediction method, by finding the historical equipment that is most similar to the equipment to be tested, and then building a prediction model based on the historical experimental data of the historical equipment, the life of the equipment to be tested is predicted by the prediction model. It can quickly predict the life of the equipment to be tested by machine learning, and can obtain the expected life of the equipment to be tested without a long life experiment. It has good timeliness and reduces testing costs, which is conducive to product updates and iterations.
[0025] In an exemplary embodiment, the preprocessing of the historical data set to obtain an optimized data set includes: performing label conversion on the life data in the historical data set to obtain a converted data set; and performing anomaly detection and missing filling on the converted data set to obtain the optimized data set.
[0026] In an exemplary embodiment, the label conversion of the life data in the historical data set to obtain the converted data set includes: finding the range interval to which the life data belongs; mapping the life data to the label value corresponding to the range interval to obtain the converted data set.
[0027] As mentioned above, the MTTF value (Mean Time To Failure, MTTF) is usually a large value, usually thousands or tens of thousands of hours, which will cause inconvenience in data processing. Therefore, this application defines different label values for different range intervals, and maps the life data to the corresponding label value based on the range interval where the life data is located, which can solve this problem well. The specific mapping rules can be seen in the following table, Table 1: After the mapping is completed, the transformed data set is obtained, and the transformed data set needs to be processed in the next step, namely: anomaly detection and missing filling.
[0028] For example, the isolation forest method can be used for anomaly detection. The idea of the isolation forest method is as follows: if there is a set of one-dimensional data that needs to be split, and the purpose is to separate the discrete points separately, then first randomly select a value X between the maximum and minimum values, and then divide the data into two groups according to less than X and greater than X. Repeat this step in these two groups of data until the data can no longer be divided. The discrete points are relatively isolated from other data, and it may take a few times to separate them. Non-discrete points are clustered with other data points, and it may take more times to separate them.
[0029] From a statistical point of view, relatively clustered points require more segmentation times, and relatively isolated points require fewer segmentation times. Isolation forest uses the number of segmentation times to measure whether a point is clustered (normal value) or isolated (outlier).
[0030] Based on this, in a specific embodiment of the present application, when the data is two-dimensional data, multiple trees can be generated and thresholds can be randomly set. As each tree grows, when a certain height is reached or each leaf node has only one point, those points with relatively shallow depth can be quickly identified, and these points are outlier points.
[0031] For details, please refer to the following sub-steps: Step 201, randomly select n sample points from the transformed data set to form subsets, and build a tree on the m subsets; Step 202, randomly select a feature and randomly select a threshold for binary splitting; Step 203, recursively build the tree in step 202 until a certain height is reached or each leaf has only one point; Step 204, define the abnormal probability according to the average depth of the m decision trees: The abnormal probability is defined based on formula 1: ; in For sample data x In by Anomaly index in iTree (isolation tree); Represents sample points x The path length, For sample x The expectation of path lengths in a batch of isolated trees; Represents the path length of each tree. Calculated by formula 2 , Formula 2: ; in, is the harmonic number, which can be estimated as , For a given number of samples When , the average path length is used to standardize the sample x The path length .
[0032] When the anomaly index is close to 1, the sample x will be judged as abnormal and the abnormal value will be removed. The specific judgment range depends on the actual situation and is not specifically limited here.
[0033] When filling missing values, a feature set is formed based on normal samples. The rows in the feature set represent each sample, and the columns represent a certain feature. When there is a missing value in a sample, a feature value is randomly extracted from the column where the missing value is located to fill the missing value.
[0034] In an exemplary embodiment, constructing historical feature data based on the optimized data set includes: The optimized data set is screened based on preset indicators to obtain screening results, wherein the preset indicators include current data indicators; and the historical characteristic data is constructed based on the screening results.
[0035] In an exemplary embodiment, the screening of the optimized data set based on a preset indicator to obtain a screening result includes: The first-order difference value of the preset index of each element in the optimized data set is obtained; if the first-order difference value of the current element is greater than the effective threshold, the current element is determined as a valid element; and the set of all the valid elements is determined as the screening result.
[0036] Exemplarily, the preset indicator can select temperature-cycle current data. The temperature-cycle experiment is an experiment to verify the impact of the simulated temperature alternating environment on the mechanical and electrical properties of electronic components. It can assess the ability of electronic components to withstand repeated temperature changes in a short period of time and the thermal matching performance between different structural materials, expose potential material defects and manufacturing quality defects of components, eliminate early failures, and improve product reliability. When the first-order difference of the temperature-cycle current data is too low, it may mean that the temperature has changed but the current change is very small, and there is a fault in the circuit, so this part of the data is not used. Exemplarily, samples with a first-order difference of the temperature-cycle current data greater than the 80mA threshold can be selected for feature construction.
[0037] In an exemplary embodiment, based on the historical feature data, obtaining a target device having a similarity with the device to be tested greater than a similarity threshold from a historical database includes: Based on the clustering algorithm and the historical feature data, a target device whose similarity with the device to be tested is greater than the similarity threshold is searched from the historical database.
[0038] Commonly used clustering algorithms include K-Means clustering algorithm, DBSCAN clustering algorithm, hierarchical clustering, etc. In this implementation, the DBSCAN clustering algorithm is taken as an example to introduce the classification process of this implementation.
[0039] The DBSCAN algorithm can match the machine that is most similar to the machine to be predicted from the historical machines. Specifically, suppose the refrigerator sample set is ,The refrigerator training set is divided by DBSCAN clustering algorithm.
[0040] The specific steps are: Given area radius Eps and the minimum number of points within the area radius to become a core object . Randomly select a sample point from the sample ,by Centered on is the radius, and the number of samples within the radius is determined. relationship, when ,but The sample within the range is the core object, otherwise it is a noise point. Then the core object is regarded as the center of the circle, and the above steps are repeated until all sample points are traversed. The object in cluster C finally found is the machine that is most similar to the machine to be predicted in this embodiment.
[0041] Among them, the number of samples The definition of is: ,when Neighborhood contains sample set Zhongyu The distance is less than or equal to The number of sub-sample sets is recorded as ; The definition of the core object is: for any sample , if its Neighborhood corresponding At least samples, that is, if ,but Is the core object.
[0042] Finally, when building a prediction model, the model training set can be constructed through the experimental data of the target machine. Specifically, the initial model for refrigerator life prediction is built based on the LightGBM algorithm, and the algorithm parameters of LightGBM are optimized through Bayesian optimization. At the same time, the processed experimental data is divided into a training set and a test set, and a 10-fold cross-validation is performed to train the initial model and obtain a prediction model.
[0043] Based on the prediction model, the life of the device under test is predicted, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test. The collected data may include refrigerator incoming data, component data, temperature cycle current, speed and other data, and the output life prediction result is the label value shown in Table 1.
[0044] In summary, the embodiment of the present application finds the historical device that is most similar to the device under test, and then builds a prediction model based on the historical experimental data of the historical device, and predicts the life of the device under test through the prediction model. It can quickly predict the life of the device under test through machine learning, and can quickly obtain the expected life of the device under test without life experiments. At the same time, compared with random inspections, it has higher accuracy, better timeliness and reduces testing costs, which is conducive to product updates and iterations.
[0045] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.
[0046] Based on the same inventive concept, the embodiment of the present application also provides a refrigerator life prediction device for implementing the refrigerator life prediction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more refrigerator life prediction device embodiments provided below can refer to the limitations of the refrigerator life prediction method above, and will not be repeated here.
[0047] In an exemplary embodiment, Figure 2 As shown, a refrigerator life prediction device 200 is provided, comprising: a preprocessing module 201, a construction module 202, an acquisition module 203 and a prediction module 204, wherein: A preprocessing module 201 is used to obtain historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; A construction module 202, configured to construct historical feature data based on the optimized data set; An acquisition module 203 is used to acquire, from a historical database, a target device whose similarity with the device to be tested is greater than a similarity threshold based on the historical feature data; The prediction module 204 is used to build a prediction model based on the experimental data of the target device, and predict the life of the device under test based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0048] In an exemplary embodiment, the preprocessing module 201 is further used for: Performing label conversion on the lifespan data in the historical data set to obtain a converted data set; Anomaly detection and missing fill are performed on the converted data set to obtain the optimized data set.
[0049] In an exemplary embodiment, the preprocessing module 201 is further used for: Find the range interval to which the life span data belongs; The life data is mapped to the label value corresponding to the range interval to obtain the converted data set.
[0050] In an exemplary embodiment, the construction module 202 is further used to: Screening the optimized data set based on preset indicators to obtain screening results, wherein the preset indicators include current data indicators; The historical feature data is constructed based on the screening result.
[0051] In an exemplary embodiment, the construction module 202 is further used to: Obtaining a first-order difference value of a preset index of each element in the optimized data set; If the first-order difference value of the current element is greater than the valid threshold, the current element is determined as a valid element; A set of all the valid elements is determined as the screening result.
[0052] In an exemplary embodiment, the acquisition module 203 is further configured to: Based on the clustering algorithm and the historical feature data, a target device whose similarity with the device to be tested is greater than the similarity threshold is searched from the historical database.
[0053] Each module in the above refrigerator life prediction device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.
[0054] The embodiment of the present application finds the historical device that is most similar to the device under test, and then builds a prediction model based on the historical experimental data of the historical device, and predicts the life of the device under test through the prediction model. It can quickly predict the life of the device under test through machine learning, and can quickly obtain the expected life of the device under test without life experiments. At the same time, compared with random inspections, it has higher accuracy, better timeliness and reduces testing costs, which is conducive to product updates and iterations.
[0055] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 3As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a method for predicting the life of a refrigerator is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.
[0056] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0057] The embodiment of the present application finds the historical device that is most similar to the device under test, and then builds a prediction model based on the historical experimental data of the historical device, and predicts the life of the device under test through the prediction model. It can quickly predict the life of the device under test through machine learning, and can quickly obtain the expected life of the device under test without life experiments. At the same time, compared with random inspections, it has higher accuracy, better timeliness and reduces testing costs, which is conducive to product updates and iterations.
[0058] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0059] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0060] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
[0061] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.
[0062] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.
[0063] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.
Claims
1. A refrigerator life prediction method, characterized in that: The method comprises: Acquire historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; constructing historical feature data based on the optimized data set; Based on the historical feature data, a target device whose similarity to the device to be tested is greater than a similarity threshold is obtained from a historical database; A prediction model is constructed based on the experimental data of the target device, and the life of the device under test is predicted based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
2. The refrigerator life prediction method according to claim 1, characterized in that: The preprocessing of the historical data set to obtain an optimized data set includes: Performing label conversion on the lifespan data in the historical data set to obtain a converted data set; Anomaly detection and missing fill are performed on the converted data set to obtain the optimized data set.
3. The refrigerator life prediction method according to claim 2, characterized in that: The label conversion of the life data in the historical data set to obtain a converted data set includes: Find the range interval to which the life span data belongs; The life data is mapped to the label value corresponding to the range interval to obtain the converted data set.
4. The refrigerator life prediction method according to claim 1, characterized in that: The constructing historical feature data based on the optimized data set includes: Screening the optimized data set based on preset indicators to obtain screening results, wherein the preset indicators include current data indicators; The historical feature data is constructed based on the screening result.
5. The refrigerator life prediction method according to claim 4, characterized in that: The step of screening the optimized data set based on the preset index to obtain the screening result includes: Obtaining a first-order difference value of a preset index of each element in the optimized data set; If the first-order difference value of the current element is greater than the valid threshold, the current element is determined as a valid element; A set of all the valid elements is determined as the screening result.
6. The refrigerator life prediction method according to claim 1, characterized in that: The step of acquiring, from a historical database, a target device whose similarity to the device to be tested is greater than a similarity threshold based on the historical feature data includes: Based on the clustering algorithm and the historical feature data, a target device whose similarity with the device to be tested is greater than the similarity threshold is searched from the historical database.
7. A refrigerator life prediction device, characterized in that: The device comprises: A preprocessing module, used to obtain historical data sets of multiple historical devices, and preprocess the historical data sets to obtain optimized data sets; A construction module, used for constructing historical feature data based on the optimized data set; An acquisition module, configured to acquire, from a historical database, a target device whose similarity to the device to be tested is greater than a similarity threshold based on the historical feature data; A prediction module is used to build a prediction model based on the experimental data of the target device, and predict the life of the device under test based on the prediction model, wherein the input parameter of the prediction model is the collected data of the device under test, and the output parameter of the prediction model is the life prediction result of the device under test.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.