Method and device for determining equipment state, cloud equipment and storage medium

Abstract

The invention relates to an equipment state determination method and apparatus, a cloud device and a storage medium. The method comprises the steps of obtaining an operation data set of a target device under each time scale in a time scale set and an environment data set of an outdoor environment where the target device is located; for each time scale in the time scale set, determining scale characteristics corresponding to the target equipment under the time scale according to the operation data set and the environment data set corresponding to the time scale, and determining a target loss degree corresponding to the target equipment under the time scale according to the environment data set corresponding to the time scale; and on the basis of scale features corresponding to the time scales in the time scale set and the target loss degree, using a state prediction model to predict the loss state of the target equipment so as to determine and obtain a target loss state corresponding to the target equipment. Reliable operation of the target equipment and the use experience of the user are ensured.

Description

Technical Field

[0001] This application relates to the field of smart home technology, and in particular to a method, apparatus, cloud device, and storage medium for determining device status. Background Technology

[0002] With the popularization of IoT technology, smart homes are gradually being integrated into daily life. To ensure the stable operation of smart homes and extend their service life, determining the status of smart homes is intuitive and important.

[0003] Currently, in order to determine the status of smart homes, a fixed threshold is usually set for the data of a certain sensor to acquire the data collected by the sensor in real time. By comparing the real-time acquired data with the fixed threshold, the status of the smart home is determined. Based on the determined status, if the smart home is found to have potential dangers, a maintenance prompt is triggered.

[0004] Because the above-mentioned methods for determining the status of smart homes use a single data source, while the actual status of smart homes is affected by a variety of factors, these methods cannot accurately determine the status of smart homes, which can easily lead to false alarms and missed alarms, thus affecting the user experience. Summary of the Invention

[0005] This application provides a method, apparatus, cloud device, and storage medium for determining the status of a device, in order to solve the problem that the determination of the status of smart homes in the prior art is unreliable and affects the user experience.

[0006] Firstly, this application provides a method for determining the state of a device, comprising: Obtain the operational dataset of the target device at each time scale in the time scale set, as well as the environmental dataset of the outdoor environment where the target device is located; For each time scale in the time scale set, based on the running dataset and the environment dataset corresponding to the time scale, the scale characteristics corresponding to the target device under the time scale are determined, and based on the environment dataset corresponding to the time scale, the target loss degree corresponding to the target device under the time scale is determined. Based on the scale characteristics and target loss degree corresponding to each time scale in the time scale set, the loss state of the target device is predicted using a state prediction model to determine the target loss state corresponding to the target device.

[0007] In an optional implementation, before performing the step of determining the target wear level of the target device at the time scale based on the environmental dataset corresponding to the time scale, the method further includes: Obtain the operation dataset corresponding to the target device at each time scale in the time scale set, wherein each operation data in the operation dataset is used to indicate the data generated by the object performing an operation on the target device; Determining the target wear level of the target device at the time scale based on the environmental dataset corresponding to the time scale includes: Based on the environmental dataset and the operational dataset corresponding to the time scale, the target wear level of the target device at the time scale is determined.

[0008] In an optional implementation, determining the target wear level of the target device at the time scale based on the environmental dataset and the operational dataset corresponding to the time scale includes: Based on the environmental dataset corresponding to the time scale, determine the initial wear level of the target device at the time scale; Based on the operational dataset, determine the target degree correction value corresponding to the initial degree of loss; The initial loss level is corrected using the target level correction value to obtain the target loss level of the target device at the time scale.

[0009] In one optional implementation, the environmental dataset corresponding to the time scale includes a subset of environmental data for multiple environmental parameters; The step of determining the initial wear level of the target device at the time scale based on the environmental dataset corresponding to the time scale includes: Determine the data statistics type corresponding to the time scale; For each environmental parameter in the environmental dataset corresponding to the time scale, based on the data statistics type, all environmental data in the environmental data subset of the environmental parameter are statistically analyzed to obtain the data statistics value corresponding to the environmental parameter, and the target score corresponding to the environmental parameter is determined based on the preset statistical value range to which the data statistics value corresponding to the environmental parameter belongs. Based on the preset weights corresponding to each of the environmental parameters and the target score, the initial degree of damage to the target device at the time scale is determined. Determining the target level correction value corresponding to the initial level of damage based on the operational dataset includes: Based on the operation dataset, determine the target frequency of the target operation performed by the object on the target device within the time scale, and determine the initial degree correction value corresponding to the initial degree of loss based on the preset frequency interval to which the target frequency belongs; The initial degree correction value is corrected using the target frequency to determine the target degree correction value corresponding to the initial degree of loss.

[0010] In an optional implementation, determining the scale characteristics of the target device at the time scale based on the runtime dataset and the environment dataset corresponding to the time scale includes: Frequency domain analysis is performed on the operational dataset corresponding to the time scale using Fourier transform to extract the operational features of the target device at the time scale from the operational dataset. Fourier transform is used to perform frequency domain analysis on the environmental dataset corresponding to the time scale, so as to extract the environmental features corresponding to the target device at the time scale from the environmental dataset. The operational characteristics and environmental characteristics corresponding to the time scale are determined as the scale characteristics of the target device under the time scale.

[0011] In one optional implementation, the state prediction model has multiple prediction layers and a fusion layer, with each prediction layer corresponding one-to-one with a time scale in the time scale set. The step of predicting the loss state of the target device using a state prediction model based on the scale characteristics corresponding to each time scale in the time scale set and the target loss degree, to determine the target loss state corresponding to the target device, includes: For each time scale in the time scale set, the scale feature corresponding to the time scale and the target loss degree are input into the prediction layer corresponding to the time scale in the state prediction model, so that the prediction layer predicts the loss state of the target device under the time scale, and obtains the prediction result corresponding to the target device under the time scale. The prediction results corresponding to all the target devices are input into the fusion layer in the state prediction model, so that the fusion layer fuses all the prediction results to obtain the target loss state corresponding to the target device.

[0012] In an optional implementation, after obtaining the target loss state, the method further includes: The target equipment is maintained according to the target wear status; Obtain the actual maintenance data recorded during the maintenance of the target device; The actual wear and tear status of the target equipment is determined from the actual maintenance data. Based on the actual loss state and the target loss state, determine the actual accuracy of the state prediction model in predicting the loss state of the target device; The current preset expansion factor is updated using the actual accuracy. Based on the updated preset expansion factor, the first association relationship is queried to obtain the time scale set and the state prediction model again. The first association relationship stores multiple sets of correspondences between the preset expansion factor, the time scale set and the state prediction model. The steps of obtaining the target device's operational dataset and the environmental dataset of the target device's outdoor environment at each time scale in the time scale set are performed using the re-obtained time scale set and the state prediction model.

[0013] Secondly, this application provides a device for determining the state of an equipment, comprising: The acquisition module is used to acquire the operational dataset of the target device at each time scale in the time scale set and the environmental dataset of the outdoor environment where the target device is located. The determination module is used to determine the scale characteristics of the target device at each time scale in the time scale set, based on the running dataset and the environment dataset corresponding to the time scale, and to determine the target loss level of the target device at the time scale based on the environment dataset corresponding to the time scale. The determining module is further configured to predict the loss state of the target device based on the scale features corresponding to the time scale in the time scale set and the target loss degree, using a state prediction model to obtain the target loss state corresponding to the target device.

[0014] Thirdly, this application provides a cloud device, including a processor and a memory, wherein the processor is configured to execute a device state determination program stored in the memory to implement the device state determination method described above.

[0015] Fourthly, this application provides a storage medium storing one or more programs that can be executed by one or more processors to implement the device state determination method described above.

[0016] Compared with the prior art, the technical solutions provided in this application have the following advantages. The method for determining the device state provided in this application includes: acquiring the operating dataset of the target device and the environmental dataset of the outdoor environment where the target device is located at each time scale in the time scale set; for each time scale in the time scale set, determining the scale characteristics of the target device at the time scale based on the operating dataset and environmental dataset corresponding to the time scale, and determining the target wear level of the target device at the time scale based on the environmental dataset corresponding to the time scale; and predicting the wear state of the target device using a state prediction model based on the scale characteristics and target wear level of the time scale in the time scale set, so as to determine the target wear state of the target device. By acquiring operational datasets of the target device and its outdoor environment datasets at multiple time scales, this application determines the scale characteristics and target wear level at each time scale based on the acquired data. These multi-scale characteristics and target wear level are then input into the state prediction model to comprehensively predict the wear state of the target device. This achieves dynamic perception of the target device's wear state, avoids the low reliability problem of using a single data threshold to determine the target device's state, improves the accuracy of determining the target device's state, and ensures the reliable operation of the target device and the user experience. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0020] Figure 1 A flowchart illustrating a method for determining device status provided in an embodiment of this application; Figure 2 A flowchart illustrating another method for determining device status provided in an embodiment of this application; Figure 3A schematic diagram of the structure of a device for determining the state of a device provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of a cloud device provided in an embodiment of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] The following disclosure provides numerous different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the scope of the invention. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed.

[0023] refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for determining device status according to an embodiment of this application. The method for determining device status according to an embodiment of this application includes the following steps: S101: Obtain the operational dataset of the target device at each time scale in the time scale set, as well as the environmental dataset of the outdoor environment where the target device is located.

[0024] In this embodiment, the target device can be understood as a smart home appliance requiring maintenance, such as a smart air conditioner, smart refrigerator, or smart washing machine. The timescale set includes multiple time scales, which are divided according to different time dimensions, such as minute, hour, day, month, and year. Its function is to analyze the operating status of the target device and the environmental changes of its outdoor environment based on multiple time scales, capturing the device's operating characteristics and environmental characteristics at different time scales, thereby providing reliable basic data for subsequent maintenance of the target device. The operating data in the operating dataset consists of data generated during the operation of the target device, while the environmental data in the environmental dataset consists of environmental data of the outdoor environment in which the target device is located during its operation.

[0025] During the operation of the target device, a lightweight data agent is deployed within it. Within the home LAN, a modified IEEE 1588 protocol is used to synchronize the lightweight data agent's clock with the gateway, ensuring that subsequent data analysis is based on the same time reference. After clock synchronization, the lightweight data agent collects operational data from the target device according to a sampling period. This operational data may include current, voltage, vibration frequency, etc., and is typically sampled at high frequencies, reflecting the instantaneous operating status of the target device. The operational data varies depending on the target device and can be configured according to specific needs. Similarly, the lightweight data agent collects environmental data from the outdoor environment according to a sampling period. This environmental data may include temperature, humidity, dust concentration, smoke concentration, etc., and is typically sampled at low frequencies. This ensures that both the collected operational and environmental data are timestamped based on the same time reference.

[0026] After obtaining the operational and environmental data, it can be encrypted and transmitted to the home gateway via a secure communication protocol (e.g., MQTT over TLS). The gateway then aggregates and uploads the data to a cloud device for storage, enabling the cloud device to perform maintenance analysis on the target device based on the uploaded operational and environmental data. It should be noted that, to reduce data volume and network bandwidth consumption, the operational data can be operational characteristic data (e.g., average current, peak current, etc.) within a sliding window.

[0027] When maintenance analysis of a target device is required on a cloud device, all operational and environmental data at each time scale can be obtained from all operational and environmental data stored on the cloud device. All operational data at each time scale is regarded as the operational dataset, and all environmental data at each time scale is regarded as the environmental dataset.

[0028] S102: For each time scale in the time scale set, determine the scale characteristics of the target device at the time scale based on the corresponding runtime dataset and environment dataset, and determine the target loss level of the target device at the time scale based on the corresponding environment dataset.

[0029] In this embodiment, after obtaining the operational dataset of the target device and the environmental dataset of the outdoor environment where the target device is located at different time scales, in order to solve the problem of time scale mismatch between the operational data (high-frequency data) and the environmental data (low-frequency data) due to different sampling frequencies, the data is time-aligned so that they can be related to each other within the same analysis window, thereby improving the accuracy of subsequent target device maintenance.

[0030] During time alignment, for each time scale in the time scale set, the running dataset at that time scale is downsampled according to various preset downsampling rates in the preset downsampling rate set to obtain candidate datasets corresponding to each preset downsampling. For each candidate dataset, a preset dynamic time warping algorithm (e.g., DTW) is used to determine the cumulative distance between the candidate dataset and the environment dataset at that time scale. The candidate dataset with the smallest cumulative distance is determined from all candidate datasets. The candidate dataset with the smallest cumulative distance is used as the running dataset. Based on the path obtained by the preset dynamic time warping algorithm, the running dataset and the environment dataset are time aligned. Then, based on the aligned running dataset and environment dataset, subsequent steps for determining scale features and target loss are performed.

[0031] As mentioned above, the preset downsampling rate set may include multiple preset downsampling rates, such as 10x downsampling rate, 100x downsampling rate, and 1000x downsampling rate. The preset downsampling rate can be set according to actual needs, which will not be elaborated in this embodiment.

[0032] After mapping the operational and environmental datasets at various implementation scales, periodic features of the target device's operation are extracted from the operational dataset, and periodic features of the outdoor environmental impact are extracted from the environmental dataset. Combining these extracted periodic features yields the scale features for that time scale, reflecting the relationship between the outdoor environment and the target device's operation at that time scale. Furthermore, using the environmental dataset at that time scale, the target wear level for the target device at that time scale is determined. Based on this target wear level, the scale features can be used to predict the wear state of the target device at that time scale. This target wear level reflects the degree of impact of the outdoor environment on the target device's wear within that time scale.

[0033] S103: Based on the scale characteristics and target loss degree corresponding to each time scale in the time scale set, the loss state of the target equipment is predicted by the state prediction model to determine the target loss state corresponding to the target equipment.

[0034] In this embodiment, the state prediction model is used to predict the actual wear and tear state of the target device. This state prediction model is a pre-trained prediction model. After obtaining the scale features and target wear level corresponding to each time scale in the time scale set, the scale features and target wear level corresponding to each time scale can be used as common inputs to the state prediction model. This allows the state prediction model to perform multi-scale fusion prediction on the scale features and target wear level corresponding to different time scales, thereby obtaining the target wear state of the target device. For example, the target wear state can be the remaining service life of the target device, or the specific representation of the target wear state can be set according to actual needs.

[0035] Once the target wear and tear status is determined, the cloud can automatically generate and trigger personalized maintenance prompts based on that status. For example, when the predicted risk of air conditioner filter clogging reaches a threshold, a "filter cleaning recommended" message will be pushed to the user via a mobile app. Furthermore, if professional maintenance is deemed necessary, the system will automatically display information on nearby authorized repair shops, their contact details, and even online appointment links through the integrated local service ecosystem, forming a closed loop from prediction to service.

[0036] This embodiment provides a method for determining the state of a device. By acquiring operational datasets of the target device and environmental datasets of its outdoor environment at multiple time scales, the method determines the scale characteristics and target wear level at each time scale based on the acquired data. These multi-scale characteristics and target wear level are then input into a state prediction model to comprehensively predict the wear state of the target device. This achieves dynamic perception of the wear state of the target device, avoids the low reliability problem of using a single data threshold to determine the state of the target device, improves the accuracy of determining the state of the target device, and ensures the reliable operation of the target device and the user experience.

[0037] refer to Figure 2 , Figure 2 This is a flowchart illustrating another method for determining device status provided in this embodiment. The method for determining device status provided in this embodiment specifically includes the following steps: S201: Obtain the target device's operational dataset, the target device's outdoor environment dataset, and the target device's corresponding operational dataset at each time scale in the time scale set.

[0038] In this embodiment, the method for obtaining the runtime dataset and environment dataset is consistent with step S101 described above, and can be referred to step S101 for details. This embodiment will not repeat the details here. Each operation data in the operation dataset is used to indicate the data generated by an object operating on the target device. The object can be understood as a user operating the target device.

[0039] Since the damage to the target device is not only related to the outdoor environment in which it is located, but also to improper operation by the object, this embodiment acquires all operational data generated by the object's operation on the target device at each preset scale within a preset scale set. All operational data acquired at each preset scale is considered as the operational dataset corresponding to the target device. Therefore, for each time scale, the target damage level of the target device is determined using the corresponding environmental dataset and operational dataset. Through this method, this embodiment acquires and analyzes the operational data generated by the object's operation on the target device, incorporating it as a key variable into the calculation of the target damage level, thus avoiding deviations in determining the damage state of the target device due to ignoring human factors.

[0040] The aforementioned timescale set and the following state prediction model can be determined in the following way: Get the current preset expansion factor; Based on the current preset expansion factor, query the first association relationship to obtain the time scale set and state prediction model corresponding to the preset expansion factor.

[0041] The preset expansion factor is used to update the time scale set and the state prediction model. The first association relationship stores the correspondence between multiple sets of preset expansion factors, time scale sets, and state prediction models. When the step of determining the equipment state needs to be executed, the current preset expansion factor is first obtained. Then, the obtained preset expansion factor is used to query the first association relationship to obtain the corresponding time scale set and state prediction model from the first association relationship based on the preset expansion factor. Then, based on the time scale set, the steps of obtaining the target equipment's operating dataset and the environmental dataset of the target equipment's outdoor environment at each time scale in the time scale set are executed. In this way, by introducing the preset expansion factor as a unified control parameter and dynamically mapping the specific time scale set and state prediction model based on the pre-established first association relationship, the flexible configuration of the time scale set and state prediction model is realized. This avoids the problem of poor adaptability of the state prediction model caused by using a fixed time scale set and improves the prediction accuracy of the subsequent state prediction model.

[0042] S202: For each time scale in the time scale set, determine the scale characteristics of the target device at the time scale based on the corresponding operational dataset and environmental dataset, and determine the target wear level of the target device at the time scale based on the corresponding environmental dataset and operational dataset.

[0043] In this embodiment, step S202 determines the scale characteristics of the target device at the time scale based on the runtime dataset and environment dataset corresponding to the time scale, including: Fourier transform is used to perform frequency domain analysis on the runtime dataset corresponding to the time scale in order to extract the runtime characteristics of the target device at the time scale from the runtime dataset. Fourier transform is used to perform frequency domain analysis on the environmental dataset corresponding to the time scale in order to extract the environmental features of the target device at the time scale from the environmental dataset. The operational and environmental characteristics corresponding to the time scale are defined as the scale characteristics of the target equipment under the time scale.

[0044] Both the running dataset and the environment dataset are time-domain signals, and they have already been time-aligned. Therefore, by using Fourier transform to perform frequency domain analysis on the above time-domain signals, the scale characteristics corresponding to the time scale can be obtained.

[0045] Specifically, Fourier transforms are performed on the operational dataset and the environmental dataset separately to calculate the power spectrum. Feature recognition is performed on the power spectrum corresponding to the operational dataset to obtain the operational characteristics of the target device, and feature recognition is performed on the power spectrum corresponding to the environmental dataset to obtain the environmental characteristics of the target device. The operational characteristics and environmental characteristics at the same time scale are concatenated to obtain the scale characteristics corresponding to that time scale. It should be noted that the specific implementation of feature extraction using Fourier transform can refer to existing technologies, and will not be elaborated here. Through the above method, this embodiment uses Fourier transform to perform frequency domain analysis on the operational dataset and environmental data to extract the operational characteristics and external environmental characteristics representing the target device at the internal time scale, and merges the two into a unified scale characteristic. This achieves the transformation from a mixed time domain signal to a clear frequency domain pattern, avoids the problems of time domain noise interference and insignificant periodic feature extraction, and improves the accuracy of predicting the target loss state.

[0046] In step S202 above, the target wear level of the target device at the time scale is determined based on the environmental dataset and operational dataset corresponding to the time scale, including: Based on the environmental dataset corresponding to the time scale, determine the initial wear level of the target device at that time scale. Based on the operational dataset, determine the target level correction value corresponding to the initial level of damage; The initial loss level is corrected using the target level correction value to obtain the target loss level of the target equipment on the time scale.

[0047] The environmental dataset comprises subsets of environmental data with multiple environmental parameters, including temperature, humidity, and salt spray concentration. Since the environmental dataset has already been time-aligned, it is structured into a matrix. Each row of this matrix represents a time point (i.e., the row includes environmental data for multiple parameters at the same time point), and each column represents an environmental parameter. Using the Eigen library, the initial wear level of the target device at that time scale is calculated. After calculating the initial wear level, all operational data in the operational dataset corresponding to the same time scale are statistically analyzed to obtain the statistical results for the operational dataset. These statistical results are then used to determine the target wear level correction value corresponding to the initial wear level. Finally, the target wear level correction value is added to the initial wear level to obtain a more accurate target wear level for the target device, ensuring the accuracy of subsequent predictions of the target device's wear state. In this embodiment, the initial wear level is first calculated based on the environmental dataset, and then the target wear level correction value is calculated based on the user's operation data. The two are then merged and corrected, thereby achieving the prediction of the target wear state of the target device through the collaborative prediction of the environment and user behavior. This avoids deviations in determining the wear state of the target device due to ignoring human factors.

[0048] In the above, based on the environmental dataset corresponding to the time scale, the initial wear level of the target device at that time scale is determined, specifically including: Determine the data statistics type corresponding to the time scale; For each environmental parameter in the environmental dataset corresponding to the time scale, based on the data statistics type, all environmental data in the environmental data subset of the environmental parameter are statistically analyzed to obtain the data statistics value corresponding to the environmental parameter, and the target score corresponding to the environmental parameter is determined based on the preset statistical value range to which the data statistics value corresponding to the environmental parameter belongs. Based on the preset weights and target scores corresponding to each environmental parameter, the initial degree of damage to the target equipment at the time scale is determined.

[0049] The data statistics type specifies the statistical method used to analyze all environmental data within a subset of environmental data. Data statistics types can include average, maximum, and minimum values, and can be set according to actual needs. Multiple time scales and their corresponding data statistics types can be pre-defined. When determining the initial wear level of a target device at a specific time scale, the above correspondence can be queried based on that time scale to obtain the corresponding data statistics type. This data statistics type can then be used for subsequent statistical calculations on various subsets of environmental data.

[0050] After obtaining the data statistical type corresponding to the data scale, for each environmental parameter in the environmental dataset corresponding to the time scale, based on the data statistical type, statistics are performed on all environmental data in the environmental data subset of the environmental parameter according to the data statistical type, thereby obtaining the data statistical value of the environmental parameter corresponding to the data statistical type. For example, when the data statistical type is average, the data statistical value is also the average value in the environmental data subset.

[0051] After obtaining the statistical data, the target score corresponding to the statistical data interval can be determined based on multiple preset statistical value intervals corresponding to pre-set environmental parameters. Furthermore, based on the correspondence between these preset statistical value intervals and preset scores, the target score corresponding to the preset statistical value interval can be determined. For example, when the statistical data is an average value, multiple average value intervals can be preset, each with a corresponding preset score. Therefore, after obtaining the statistical data, the average value interval to which the statistical data belongs can be determined, and the preset score set for that average value interval can be determined based on the determined average value interval. This determined preset score can then be used as the target score corresponding to the environmental parameters.

[0052] After obtaining the target scores corresponding to each environmental parameter, the preset weights corresponding to each environmental parameter can be determined based on the pre-set correspondence between multiple environmental parameters and preset weights. Then, by multiplying the target scores corresponding to all environmental parameters by their corresponding preset weights and summing the results, the initial wear level of the target device at the time scale can be obtained. The initial wear level is a quantified value. It should be noted that the initial wear level is determined by processing the matrix using the Eigen library based on the above determination method. Through the above method, this embodiment defines the data statistical type for the time scale, maps the data statistical values ​​obtained from each environmental parameter to a unified target score, and then performs a weighted calculation based on preset weights reflecting the importance of the environmental parameters. This achieves the integration of multi-dimensional heterogeneous environmental data into a precise quantified initial wear level, ensuring the accuracy of subsequent determination of the target wear level based on the initial wear level.

[0053] In the above, based on the operational dataset, the target level correction value corresponding to the initial level of damage is determined, including: Based on the operation dataset, determine the target frequency of the target operation performed by the object on the target device within the time scale, and determine the initial degree correction value corresponding to the initial degree of loss based on the preset frequency interval to which the target frequency belongs; The initial level correction value is adjusted using the target frequency to determine the target level correction value corresponding to the initial level of loss.

[0054] After obtaining the operation dataset, the target frequency of the target operation performed by the object on the target device within a time scale is statistically analyzed. This target operation can be a power-on / off operation, and the target frequency can be understood as the number of times the object powers on and off the target device, reflecting the frequency with which the object performs the target operation. The target frequency is determined from multiple pre-set frequency intervals, and the correspondence between these intervals and the initial severity correction value is established. The initial severity correction value is obtained by querying this correspondence using the determined frequency interval. Since the more severe the operation, the greater the penalty among similar adverse operations, the target frequency is used to correct the initial severity correction value after it is obtained, resulting in a more accurate target severity correction value. When correcting the initial severity correction value using the target frequency, a correction coefficient is determined based on the base frequency and target frequency corresponding to the pre-set frequency interval to which the target frequency belongs. Multiplying the initial severity correction value by the correction coefficient yields the target severity correction value corresponding to the initial damage level. The correction factor is equal to the ratio between the target frequency and the reference frequency. The reference frequency corresponding to the preset frequency range can be set according to actual needs, and is usually the median or upper limit of the preset frequency range. In this embodiment, by statistically analyzing the target frequency of user operations, an initial degree correction value is determined based on the target frequency. This initial degree correction value is then refined using the target frequency, achieving precise quantification of user operation behavior into a target degree correction value. Consequently, a more accurate target loss level can be obtained using the target degree correction value, ensuring the accuracy of subsequent determination of the target loss state.

[0055] S203: Based on the scale characteristics and target loss degree corresponding to each preset scale in the time scale set, the loss state of the target equipment is predicted by the state prediction model to determine the target loss state corresponding to the target equipment.

[0056] In this embodiment, the state prediction model has multiple prediction layers and a fusion layer, with each prediction layer corresponding one-to-one with a time scale in the time scale set. Step S203 specifically includes: For each time scale in the time scale set, the scale features and target loss level corresponding to the time scale are input into the prediction layer corresponding to the time scale in the state prediction model, so that the prediction layer can predict the loss state of the target equipment under the time scale and obtain the prediction result corresponding to the target equipment under the time scale. The prediction results for all target devices are input into the fusion layer of the state prediction model so that the fusion layer can fuse all prediction results to obtain the target loss state for the target device.

[0057] In this context, when the timescale set includes three timescales, the state prediction model comprises three prediction layers: a micro-prediction layer, a meso-prediction layer, and a macro-prediction layer. The micro-prediction layer can employ a one-dimensional convolutional neural network (1D-CNN), using small convolutional kernels to scan minute-level data and capture transient anomalies and short-term waveform patterns. The meso-prediction layer can employ a Long Short-Term Memory (LSTM) network or dilated convolutions to process daily-level data and learn the daily and weekly cycles of device operation. The macro-prediction layer can employ a Transformer encoder to analyze monthly and yearly data, capturing seasonal decay and long-term aging trends. Of course, the correspondence between the timescales in the timescale set and the prediction layers in the state prediction model can be set according to actual needs; this embodiment does not impose any limitations.

[0058] After obtaining the scale features and target loss levels corresponding to each time scale in the time scale set, for each time scale, the corresponding scale features and target loss levels are input into the prediction layer of the state prediction model. This allows the prediction layer to predict the loss state of the target device at that time scale based on the input scale features and target loss levels, and the resulting value is considered the prediction result for the target device at that time scale. After all prediction layers in the state prediction model output their prediction results, all prediction results are aggregated and input into the fusion layer. The fusion layer uses an attention mechanism to fuse all prediction results to output the target loss state of the target device, which may include the risk and remaining service life of the target device. Through this method, this embodiment achieves multi-scale collaborative analysis of the loss state of the target device by configuring a dedicated prediction layer for each time scale in the time scale set for preliminary prediction and using the fusion layer to adaptively weight and fuse all preliminary prediction results. This avoids feature interference and scale limitations when using a single prediction model to process multi-scale data, thus improving the accuracy of determining the target loss state.

[0059] After obtaining the target loss state, the method for determining the device state provided in this embodiment further includes the following steps: Maintenance is performed on the target equipment based on its wear and tear status. Obtain the actual maintenance data recorded on the target equipment; Determine the actual wear and tear status of the target equipment from actual maintenance data; Based on the actual loss state and the target loss state, determine the actual accuracy of the state prediction model in predicting the loss state of the target equipment. Update the current preset expansion factor using the actual accuracy; Based on the updated preset expansion factor, the first association relationship is queried to obtain the time scale set and state prediction model again. The first association relationship stores the correspondence between multiple sets of preset expansion factors, time scale sets and state prediction models. The steps of obtaining the operational dataset of the target device at each time scale in the time scale set and the environmental dataset of the outdoor environment where the target device is located are performed by using the re-obtained time scale set and the state prediction model.

[0060] In this process, after maintenance personnel perform maintenance on the target equipment based on the target wear status, they record the actual maintenance data of the target equipment in the cloud. The cloud can then analyze this data to obtain an actual wear status consistent with the output dimension of the status prediction model. For example, if the actual maintenance data shows that the air conditioning filter is 90% clogged, this data can be quantified to obtain an actual wear status of 0.9. It should be noted that a pre-trained machine learning model can be used to input the actual maintenance data, allowing the model to output the corresponding target wear status for the target equipment.

[0061] After obtaining the actual loss state, the absolute difference between the actual loss state and the target loss state can be calculated. Subtracting this absolute difference from 1 yields the actual accuracy of the state prediction model in predicting the loss state of the target equipment. The target accuracy corresponding to the state prediction model is then determined; this target accuracy is the accuracy that the pre-set state prediction model can achieve when predicting the loss state. The difference between the actual accuracy and the target accuracy is used to update the current preset expansion factor to obtain the updated preset expansion factor. The update formula for the preset expansion factor is as follows:

[0062] In the above, This represents the updated preset expansion factor. This represents the preset expansion factor before adjustment. This indicates the preset learning rate. Indicates actual accuracy. Indicates the target accuracy.

[0063] After obtaining the updated preset expansion factor, the first association relationship is queried using the updated target expansion factor to re-obtain the time scale set and state prediction model. The original time scale set and state prediction model are replaced with the re-obtained time scale set and state prediction model, respectively, thus completing the update of the time scale set and state prediction model. Then, based on the updated time scale set and state prediction model, the above-mentioned step S201 is executed. In this way, this embodiment compares the actual loss state obtained after maintenance with the target loss state output by the state prediction model to calculate the actual accuracy rate. This is used as a feedback signal to dynamically update the current preset expansion factor, and based on the updated preset expansion factor, the corresponding time scale set and state prediction model are dynamically updated to improve the prediction accuracy of the target loss state corresponding to the subsequent target equipment.

[0064] This embodiment provides a method for determining the state of a device. By acquiring operational datasets of the target device and environmental datasets of its outdoor environment at multiple time scales, the method determines the scale characteristics and target wear level at each time scale based on the acquired data. These multi-scale characteristics and target wear level are then input into a state prediction model to comprehensively predict the wear state of the target device. This achieves dynamic perception of the wear state of the target device, avoids the low reliability problem of using a single data threshold to determine the state of the target device, improves the accuracy of determining the state of the target device, and ensures the reliable operation of the target device and the user experience.

[0065] refer to Figure 3 , Figure 3 This is a schematic diagram of a device for determining device status provided in an embodiment of this application. The device for determining device status provided in this application includes an acquisition module 10 and a determination module 20. The acquisition module 10 is used to acquire the operational dataset of the target device and the environmental dataset of the outdoor environment where the target device is located at each time scale in the time scale set. The determination module 20 is used to determine the scale characteristics corresponding to the target device at each time scale in the time scale set, based on the operational dataset and the environmental dataset corresponding to the time scale, and to determine the target wear level corresponding to the target device at the time scale based on the environmental dataset corresponding to the time scale. The determination module 20 is also used to predict the wear state of the target device using a status prediction model based on the scale characteristics and the target wear level corresponding to the time scale in the time scale set, so as to obtain the target wear state corresponding to the target device.

[0066] In this embodiment, the acquisition module 10 is further configured to: Obtain the operation dataset corresponding to the target device at each of the time scales in the time scale set, wherein each operation data in the operation dataset is used to indicate the data generated by the object performing operations on the target device.

[0067] In this embodiment, the determining module 20 is further configured to: Based on the environmental dataset and the operational dataset corresponding to the time scale, the target wear level of the target device at the time scale is determined.

[0068] In this embodiment, the determining module 20 is further configured to: Based on the environmental dataset corresponding to the time scale, determine the initial wear level of the target device at the time scale; Based on the operational dataset, determine the target degree correction value corresponding to the initial degree of loss; The initial loss level is corrected using the target level correction value to obtain the target loss level of the target device at the time scale.

[0069] In this embodiment, the environmental dataset corresponding to the time scale includes a subset of environmental data for multiple environmental parameters, and the determining module 20 is further configured to: Determine the data statistics type corresponding to the time scale; For each environmental parameter in the environmental dataset corresponding to the time scale, based on the data statistics type, all environmental data in the environmental data subset of the environmental parameter are statistically analyzed to obtain the data statistics value corresponding to the environmental parameter, and the target score corresponding to the environmental parameter is determined based on the preset statistical value range to which the data statistics value corresponding to the environmental parameter belongs. Based on the preset weights corresponding to each of the environmental parameters and the target score, the initial degree of damage to the target device at the time scale is determined.

[0070] In this embodiment, the determining module 20 is further configured to: Based on the operation dataset, determine the target frequency of the target operation performed by the object on the target device within the time scale, and determine the initial degree correction value corresponding to the initial degree of loss based on the preset frequency interval to which the target frequency belongs; The initial degree correction value is corrected using the target frequency to determine the target degree correction value corresponding to the initial degree of loss.

[0071] In this embodiment, the determining module 20 is further configured to: Frequency domain analysis is performed on the operational dataset corresponding to the time scale using Fourier transform to extract the operational features of the target device at the time scale from the operational dataset. Fourier transform is used to perform frequency domain analysis on the environmental dataset corresponding to the time scale, so as to extract the environmental features corresponding to the target device at the time scale from the environmental dataset. The operational characteristics and environmental characteristics corresponding to the time scale are determined as the scale characteristics of the target device under the time scale.

[0072] In this embodiment, the state prediction model has multiple prediction layers and a fusion layer, with each prediction layer corresponding one-to-one with a time scale in the time scale set; the prediction module 30 is further configured to: For each time scale in the time scale set, the scale feature corresponding to the time scale and the target loss degree are input into the prediction layer corresponding to the time scale in the state prediction model, so that the prediction layer predicts the loss state of the target device under the time scale, and obtains the prediction result corresponding to the target device under the time scale. The prediction results corresponding to all the target devices are input into the fusion layer in the state prediction model, so that the fusion layer fuses all the prediction results to obtain the target loss state corresponding to the target device.

[0073] In this embodiment, the acquisition module 10 is further configured to: After obtaining the target wear status, the target equipment is maintained according to the target wear status; Obtain the actual maintenance data recorded during the maintenance of the target device; The actual wear and tear status of the target equipment is determined from the actual maintenance data. Based on the actual loss state and the target loss state, determine the actual accuracy of the state prediction model in predicting the loss state of the target device; The current preset expansion factor is updated using the actual accuracy. Based on the updated preset expansion factor, the first association relationship is queried to obtain the time scale set and the state prediction model again. The first association relationship stores multiple sets of correspondences between the preset expansion factor, the time scale set and the state prediction model. By re-obtaining the timescale set and the state prediction model, the operational dataset of the target device at each timescale in the timescale set and the environmental dataset of the outdoor environment where the target device is located are obtained.

[0074] This embodiment provides a device for determining the state of a device. By acquiring operational datasets of the target device and environmental datasets of its outdoor environment at multiple time scales, the device determines the scale characteristics and target wear level at each time scale based on the acquired data. These multi-scale characteristics and target wear level are then input into a state prediction model to comprehensively predict the wear state of the target device. This achieves dynamic perception of the wear state of the target device, avoids the low reliability problem of determining the state of the target device using a single data threshold, improves the accuracy of determining the state of the target device, and ensures the reliable operation of the target device and the user experience.

[0075] Figure 4 This is a schematic diagram of the structure of a cloud device provided in an embodiment of this application. Figure 4 The cloud device 400 shown includes: at least one processor 401, memory 602, at least one network interface 404, and other user interfaces 403. The various components in the cloud device 400 are coupled together via a bus system 405. It is understood that the bus system 405 is used to implement communication between these components. In addition to a data bus, the bus system 405 also includes a power bus, a control bus, and a status signal bus. However, for clarity, ... Figure 4 The general designated all buses as Bus System 405.

[0076] The user interface 403 may include a display, keyboard, or clicking device (e.g., mouse, trackball, touchpad, or touchscreen).

[0077] It is understood that the memory 402 in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DRRAM). The memory 402 described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0078] In some implementations, memory 402 stores elements, executable units or data structures, or subsets thereof, or extended sets thereof: operating system 4021 and application program 4022.

[0079] The operating system 4021 includes various system programs, such as the framework layer, core library layer, and driver layer, used to implement various basic business functions and handle hardware-based tasks. The application program 4022 includes various applications, such as a media player and a browser, used to implement various application functions. The program implementing the method of this embodiment can be included in the application program 4022.

[0080] In this embodiment of the invention, the processor 401 executes the method steps provided in each method embodiment by calling the program or instructions stored in the memory 402, specifically the program or instructions stored in the application program 4022.

[0081] The methods disclosed in the above embodiments of the present invention can be applied to processor 401, or implemented by processor 401. Processor 401 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in processor 401 or by instructions in the form of software. The processor 401 may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of the present invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software units in the decoding processor. The software units may be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 402. Processor 401 reads the information in memory 402 and, in conjunction with its hardware, completes the steps of the above method.

[0082] It is understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described herein, or combinations thereof.

[0083] For software implementation, the techniques described herein can be implemented by units that perform the functions described herein. The software code can be stored in memory and executed by a processor. The memory can be implemented in the processor or external to the processor.

[0084] The cloud device provided in this embodiment can be, for example, as follows: Figure 4 The cloud device shown can perform actions such as Figure 1 and Figure 2 All steps of the method for determining the status of equipment in the middle, thereby realizing Figure 1 and Figure 2 For details on the technical effects of the method for determining the status of the equipment shown, please refer to [link / reference needed]. Figure 1 and Figure 2 The relevant descriptions are presented concisely and will not be elaborated upon here.

[0085] This invention also provides a storage medium (computer-readable storage medium). This storage medium stores one or more programs. The storage medium may include volatile memory, such as random access memory; it may also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state drive; and it may also include combinations of the above types of memory.

[0086] When one or more programs in the storage medium can be executed by one or more processors to implement the above-described method for determining the device state executed on the device state determination side.

[0087] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0088] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0089] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for determining the state of equipment, characterized in that, include: Obtain the operational dataset of the target device at each time scale in the time scale set, as well as the environmental dataset of the outdoor environment where the target device is located; For each time scale in the time scale set, based on the running dataset and the environment dataset corresponding to the time scale, the scale characteristics corresponding to the target device under the time scale are determined, and based on the environment dataset corresponding to the time scale, the target loss degree corresponding to the target device under the time scale is determined. Based on the scale characteristics and target loss degree corresponding to each time scale in the time scale set, the loss state of the target device is predicted using a state prediction model to determine the target loss state corresponding to the target device.

2. The method according to claim 1, characterized in that, Before performing the step of determining the target wear level of the target device at the time scale based on the environmental dataset corresponding to the time scale, the method further includes: Obtain the operation dataset corresponding to the target device at each time scale in the time scale set, wherein each operation data in the operation dataset is used to indicate the data generated by the object performing an operation on the target device; The step of determining the target wear level of the target device at the time scale based on the environmental dataset corresponding to the time scale includes: Based on the environmental dataset and the operational dataset corresponding to the time scale, the target wear level of the target device at the time scale is determined.

3. The method according to claim 2, characterized in that, The step of determining the target wear level of the target device at the time scale based on the environmental dataset and the operational dataset corresponding to the time scale includes: Based on the environmental dataset corresponding to the time scale, determine the initial wear level of the target device at the time scale; Based on the operational dataset, determine the target degree correction value corresponding to the initial degree of loss; The initial loss level is corrected using the target level correction value to obtain the target loss level of the target device at the time scale.

4. The method according to claim 3, characterized in that, The environmental dataset corresponding to the time scale includes a subset of environmental data with multiple environmental parameters. The step of determining the initial wear level of the target device at the time scale based on the environmental dataset corresponding to the time scale includes: Determine the data statistics type corresponding to the time scale; For each environmental parameter in the environmental dataset corresponding to the time scale, based on the data statistics type, all environmental data in the environmental data subset of the environmental parameter are statistically analyzed to obtain the data statistics value corresponding to the environmental parameter, and the target score corresponding to the environmental parameter is determined based on the preset statistical value range to which the data statistics value corresponding to the environmental parameter belongs. Based on the preset weights corresponding to each of the environmental parameters and the target score, the initial degree of damage to the target device at the time scale is determined. Determining the target level correction value corresponding to the initial level of damage based on the operational dataset includes: Based on the operation dataset, determine the target frequency of the target operation performed by the object on the target device within the time scale, and determine the initial degree correction value corresponding to the initial degree of loss based on the preset frequency interval to which the target frequency belongs; The initial degree correction value is corrected using the target frequency to determine the target degree correction value corresponding to the initial degree of loss.

5. The method according to claim 1, characterized in that, The step of determining the scale features corresponding to the target device at the time scale based on the running dataset and the environment dataset corresponding to the time scale includes: Fourier transform is used to perform frequency domain analysis on the running dataset corresponding to the time scale, so as to extract the running features of the target device at the time scale from the running dataset; Fourier transform is used to perform frequency domain analysis on the environmental dataset corresponding to the time scale, so as to extract the environmental features corresponding to the target device at the time scale from the environmental dataset. The operational characteristics and environmental characteristics corresponding to the time scale are determined as the scale characteristics of the target device under the time scale.

6. The method according to claim 1, characterized in that, The state prediction model has multiple prediction layers and a fusion layer, and the multiple prediction layers correspond one-to-one with the time scales in the time scale set. The step of predicting the loss state of the target device using a state prediction model based on the scale characteristics corresponding to each time scale in the time scale set and the target loss degree, to determine the target loss state corresponding to the target device, includes: For each time scale in the time scale set, the scale feature corresponding to the time scale and the target loss degree are input into the prediction layer corresponding to the time scale in the state prediction model, so that the prediction layer predicts the loss state of the target device under the time scale, and obtains the prediction result corresponding to the target device under the time scale. The prediction results corresponding to all the target devices are input into the fusion layer in the state prediction model, so that the fusion layer fuses all the prediction results to obtain the target loss state corresponding to the target device.

7. The method according to claim 1, characterized in that, After obtaining the target loss state, the method further includes: The target equipment is maintained according to the target wear status; Obtain the actual maintenance data recorded during the maintenance of the target device; The actual wear and tear status of the target equipment is determined from the actual maintenance data. Based on the actual loss state and the target loss state, determine the actual accuracy of the state prediction model in predicting the loss state of the target device; The current preset expansion factor is updated using the actual accuracy. Based on the updated preset expansion factor, the first association relationship is queried to obtain the time scale set and the state prediction model again. The first association relationship stores multiple sets of correspondences between the preset expansion factor, the time scale set and the state prediction model. The steps of obtaining the target device's operational dataset and the environmental dataset of the target device's outdoor environment at each time scale in the time scale set are performed using the re-obtained time scale set and the state prediction model.

8. A device for determining the state of an equipment, characterized in that, include: The acquisition module is used to acquire the operational dataset of the target device at each time scale in the time scale set and the environmental dataset of the outdoor environment where the target device is located. The determination module is used to determine the scale characteristics of the target device at each time scale in the time scale set, based on the running dataset and the environment dataset corresponding to the time scale, and to determine the target loss level of the target device at the time scale based on the environment dataset corresponding to the time scale. The determining module is further configured to predict the loss state of the target device based on the scale features corresponding to the time scale in the time scale set and the target loss degree, using a state prediction model, so as to determine the target loss state corresponding to the target device.

9. A cloud device, characterized in that, include: A processor and a memory, the processor being configured to execute a device state determination program stored in the memory to implement the device state determination method according to any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores one or more programs, which can be executed by one or more processors to implement the method for determining the device state according to any one of claims 1 to 7.

Images