Power transmission system and its condition monitoring method

Abstract

This application discloses a power transmission system and its condition monitoring method. The method includes: acquiring real-time monitoring data of at least one target during the process of a power transmission vehicle supplying power to a graphitization furnace, wherein the at least one target is used to characterize at least one monitored component in the power transmission system; acquiring key state elements and element requirement information corresponding to each of the at least one target; obtaining real-time element information of each key state element through target state analysis based on at least each key state element and each real-time monitoring data; detecting whether there is abnormal monitoring data in the real-time monitoring data based on the differences between the element requirement information and the real-time element information corresponding to each of the key state elements; and determining the abnormal state information of the power transmission system based on the abnormal monitoring data when abnormal monitoring data is detected, thereby enabling accurate monitoring of the abnormal state of the power transmission system and improving equipment safety.

Description

Technical Field

[0001] This application relates to the field of equipment monitoring technology, and in particular to a power transmission system and its condition monitoring method. Background Technology

[0002] Graphitization furnaces use high temperatures to convert carbon materials into graphite, which can then be used to produce battery anode materials. When a graphitization furnace is operating, it typically requires a large amount of electrical energy to maintain the high temperatures and ensure normal production.

[0003] However, the large amount of electrical energy transmitted and the continuous high temperature can easily cause the graphitization furnace and / or the equipment supplying it to be in an unstable / abnormal state, thereby affecting equipment safety and causing potential safety hazards. Summary of the Invention

[0004] To address the aforementioned technical problems, this application proposes a power transmission system and its status monitoring method, which can accurately monitor abnormal states of the power transmission system and improve equipment safety.

[0005] In a first aspect, embodiments of this application provide a method for monitoring the status of a power transmission system, the power transmission system including a graphitization furnace and a power transmission vehicle, the method comprising:

[0006] During the process of the power transmission vehicle supplying power to the graphitization furnace, real-time monitoring data of at least one target is acquired, wherein the at least one target is used to characterize at least one monitored component in the power transmission system;

[0007] Obtain the key state elements corresponding to each of the at least one target and the element requirement information of the key state elements;

[0008] Based at least on each of the key state elements and each of the real-time monitoring data, real-time element information of each of the key state elements is obtained through target state analysis;

[0009] Based on the differences between the element requirement information and the real-time element information corresponding to each of the key state elements, the presence of abnormal monitoring data is detected in the real-time monitoring data.

[0010] If the abnormal monitoring data is detected, the abnormal status information of the power transmission system is determined based on the abnormal monitoring data.

[0011] Optionally, the at least one target is arranged in a preset order, which is adapted to indicate the electrical signal transmission order between the at least one monitored component;

[0012] The step of obtaining real-time element information for each of the key state elements through target state analysis, based at least on each of the key state elements and each of the real-time monitoring data, includes:

[0013] Extract the corresponding real-time monitoring data features from each of the real-time monitoring data;

[0014] For each target, based at least on the real-time monitoring data features corresponding to it and the real-time monitoring data features corresponding to the targets preceding it, an artificial intelligence model is invoked to perform target state analysis, generate first element information corresponding to the target, and decompose the first element information to obtain real-time element information of each key state element corresponding to the target. The targets preceding it refer to the targets whose order in the preset order is earlier than the target.

[0015] Optionally, the artificial intelligence model includes M feature interaction layers and M feature attention layers, where M is an integer greater than 1;

[0016] The process involves, based at least on the corresponding real-time monitoring data features and the real-time monitoring data features corresponding to targets preceding it, invoking an artificial intelligence model to perform target state analysis and generate the first element information corresponding to the target, including:

[0017] Based on the real-time monitoring data features corresponding to the target preceding the target, the first feature interaction layer is invoked to generate the first reference feature;

[0018] Based at least on the first reference feature and the corresponding real-time monitoring data feature, the first feature attention layer is invoked to generate the first attention feature;

[0019] Based at least on the m-th reference feature and its corresponding real-time monitoring data feature, the m-th feature attention layer is invoked to generate the m-th attention feature;

[0020] in,

[0021] m is an integer greater than 1 and not greater than M, and the m-th reference feature is generated based on the (m-1)-th attention feature by calling the m-th feature interaction layer;

[0022] The first element information corresponding to this target is generated based on the Mth attention feature.

[0023] Optionally, for each of the stated objectives, the method further includes:

[0024] Determine the relative relationship between the monitored component corresponding to the target and the monitored component corresponding to each of the targets preceding the target, wherein the relative relationship is determined at least by the electrical signal transmission sequence between the two corresponding monitored components;

[0025] in,

[0026] The step of generating a first attention feature by calling the first feature attention layer based at least on the first reference feature and the corresponding real-time monitoring data feature includes: generating a first attention feature by calling the first feature attention layer based on the relative relationship, the first reference feature and the corresponding real-time monitoring data feature;

[0027] And / or,

[0028] The step of generating the mth attention feature by calling the mth feature attention layer based at least on the mth reference feature and the corresponding real-time monitoring data feature includes: generating the mth attention feature by calling the mth feature attention layer based on the relative relationship, the mth reference feature and the corresponding real-time monitoring data feature.

[0029] Optionally, the step of detecting whether there is abnormal monitoring data in the real-time monitoring data based on the difference between the element requirement information corresponding to each of the key state elements and the real-time element information includes:

[0030] For each of the key state elements, a first information feature of the corresponding element requirement information and a second information feature of the corresponding real-time element information are determined, and the similarity between the first information feature and the second information feature is determined, wherein the similarity is used to characterize the difference.

[0031] Based on the similarity, the system detects whether there is any abnormal monitoring data in the real-time monitoring data.

[0032] Optionally, detecting the presence of abnormal monitoring data in the real-time monitoring data based on the similarity includes:

[0033] Based on the similarity, identify the elements to be identified from all the key state elements whose similarity is lower than the similarity threshold;

[0034] Extract the data of the elements to be identified that match the elements to be identified from the real-time monitoring data;

[0035] The system detects whether there is any abnormal monitoring data in the data of the elements to be identified.

[0036] Optionally, the data of the element to be identified refers to the real-time monitoring data of the target corresponding to the element to be identified.

[0037] Optionally, the at least one monitored component includes at least one first component of the power transmission vehicle and / or at least one second component of the graphitization furnace;

[0038] At least one of the following first components of the power transmission vehicle includes at least one of the following: the output terminal of the power transmission vehicle, a conductive cable, and an electrode clamp;

[0039] At least one second component of the graphitization furnace includes at least one of the following: a graphite electrode of the graphitization furnace, or a conductive heating element inside the furnace.

[0040] Optionally, the at least one monitored component includes at least one first component of the power transmission vehicle and at least one second component of the graphitization furnace. The at least one first component of the power transmission vehicle includes the output end of the power transmission vehicle, a conductive cable, and an electrode clamp. The at least one second component of the graphitization furnace includes the graphite electrode of the graphitization furnace and a conductive heating element inside the furnace.

[0041] The electrical signal transmission sequence between at least one monitored component includes a preset electrical signal transmission sequence, and the order of each component corresponding to the preset electrical signal transmission sequence is as follows: the output terminal, the conductive cable, the electrode clamp, the graphite electrode, and the conductive heating element inside the furnace.

[0042] Secondly, embodiments of this application provide a power transmission system, which includes monitoring equipment, a graphitization furnace, and a power transmission vehicle. The monitoring equipment is configured to:

[0043] During the process of the power transmission vehicle supplying power to the graphitization furnace, real-time monitoring data of at least one target is acquired, wherein the at least one target is used to characterize at least one monitored component in the power transmission system;

[0044] Obtain the key state elements corresponding to each of the at least one target and the element requirement information of the key state elements;

[0045] Based at least on each of the key state elements and each of the real-time monitoring data, real-time element information of each of the key state elements is obtained through target state analysis;

[0046] Based on the differences between the element requirement information and the real-time element information corresponding to each of the key state elements, the presence of abnormal monitoring data is detected in the real-time monitoring data.

[0047] If the abnormal monitoring data is detected, the abnormal status information of the power transmission system is determined based on the abnormal monitoring data.

[0048] In summary, the embodiments of this application have at least the following beneficial effects:

[0049] In this embodiment of the application, during the process of the power transmission vehicle supplying power to the graphitization furnace, real-time monitoring data of at least one target is acquired, wherein the at least one target is used to characterize at least one monitored component in the power transmission system; key state elements corresponding to each of the at least one target and element requirement information of the key state elements are acquired; based at least on each key state element and each of the real-time monitoring data, real-time element information of each key state element is obtained through target state analysis; based on the differences between the element requirement information and the real-time element information corresponding to each of the key state elements, abnormal monitoring data is detected in the real-time monitoring data; if abnormal monitoring data is detected, abnormal state information of the power transmission system is determined based on the abnormal monitoring data. This allows for the initial screening of real-time monitoring data by first obtaining real-time element information of different key state elements through target state analysis, then comparing the differences between the real-time element information and the corresponding element requirement information, and finally analyzing the detected abnormal monitoring data to obtain the corresponding abnormal state information. This enables accurate identification of abnormal states in the initially screened data, thereby improving equipment safety by accurately monitoring the abnormal states of the power transmission system. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the status monitoring method for a power transmission system provided in an embodiment of this application;

[0051] Figure 2 This is a schematic diagram of the power transmission system provided in an embodiment of this application;

[0052] Figure 3 This is a schematic diagram of the structure of the computer device provided in the embodiments of this application. Detailed Implementation

[0053] 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 / examples are only a part of the embodiments / examples of this application, and not all of the embodiments / examples. Based on the embodiments / examples in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0054] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "multiple" means two or more. In the description of this application, the term "comprising" and its variations are open-ended, meaning "including but not limited to." The term "based on" means "at least partially based on." The term "according to" means "at least partially according to." The term "one embodiment / example" means "at least one embodiment / example"; the term "another embodiment / example" means "at least one additional embodiment / example"; the term "some embodiments / examples" means "at least some embodiments / examples."

[0055] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0056] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this application is for the purpose of describing specific embodiments only and is not intended to limit the application. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0057] Firstly, see [the following] Figure 1 The diagram shows a flowchart of a power transmission system status monitoring method provided in an embodiment of this application. The method can be applied to monitoring equipment (such as a computer device, an industrial control computer, etc.). The power transmission system includes a graphitization furnace and a power transmission vehicle. The method includes steps S101-S105, as detailed below.

[0058] S101, during the process of the power transmission vehicle supplying power to the graphitization furnace, real-time monitoring data of at least one target are acquired, wherein the at least one target is used to characterize at least one monitored component in the power transmission system.

[0059] In some examples, the aforementioned power supply vehicle may refer to equipment capable of supplying power to the aforementioned graphitization furnace, the bottom of which may be equipped with multiple moving mechanisms to enable the equipment to function as a self-moving power supply vehicle.

[0060] In some examples, the aforementioned at least one monitored component can refer to any one or more components in the power transmission system that can be detected by corresponding sensors. In other words, the sensor corresponding to each monitored component can be used to detect the monitored component to generate real-time monitoring data of the corresponding target (e.g., the real-time monitoring data can be obtained by preprocessing the detected data). It is understood that the types and / or numbers of sensors corresponding to different monitored components can be different, so as to set the corresponding sensors to detect the monitored component according to the characteristics of each monitored component, so that the corresponding real-time monitoring data can more accurately reflect the real-time status of the corresponding monitored component. For example, the types of sensors mentioned above can include at least one of the following: temperature sensor, current sensor, voltage sensor, vibration sensor.

[0061] It is understood that at least one of the above targets corresponds one-to-one with at least one of the above monitored components; in other words, each target is used to characterize a corresponding monitored component.

[0062] S102, obtain the key state elements corresponding to each of the at least one target and the element requirement information of the key state elements.

[0063] In some examples, each target (or each monitored component) can correspond to one or more key state elements, and each key state element can correspond to element requirement information for describing the ideal parameter value of the key state element under various operating conditions.

[0064] S103, based at least on each of the key state elements and each of the real-time monitoring data, real-time element information of each of the key state elements is obtained through target state analysis.

[0065] In some examples, a pre-trained target state analysis model can be used to determine the real-time element information of each key state element based on the key state elements and the real-time monitoring data. This target state analysis model can be a trained model capable of predicting each key state element and its corresponding real-time monitoring data as model input and the corresponding real-time element information as model output. During training, sample key state elements and their corresponding sample monitoring data can be used as sample data (this sample data also carries the expected corresponding element information label, which represents the corresponding expected element information). The predicted element information generated by the model based on this sample data is obtained. Based on the difference between the predicted element information and the expected element information, a general loss function is used to calculate the loss value. A general training algorithm (e.g., gradient descent) is then used to train the model based on this loss value, so that the trained model possesses the aforementioned capabilities. It is easy to understand that the sample data in this embodiment can be experimental data obtained in advance through multiple corresponding experiments.

[0066] As can be seen from the above examples, this application has provided at least one embodiment of target state analysis that can be completed by using only key state elements and real-time monitoring data as a basis. In other words, in the relevant embodiments of this application, real-time element information of each key state element can be obtained through target state analysis based on each key state element and each real-time monitoring data.

[0067] S104, based on the difference between the element requirement information corresponding to each of the key state elements and the real-time element information, detect whether there is abnormal monitoring data in the real-time monitoring data.

[0068] In some examples, the feature requirement information corresponding to the same key state feature can be compared with the real-time feature information so that the comparison result represents the corresponding differences.

[0069] In some examples, the relevant data of each key state element whose difference in the real-time monitoring data is greater than a preset difference threshold can be directly used as anomaly monitoring data. In other words, in this embodiment, if the difference corresponding to a certain key state element is greater than a preset difference threshold, it can be considered that anomaly monitoring data can be detected in the real-time monitoring data, that is, the relevant data of the key state element whose difference is greater than the preset difference threshold.

[0070] S105, if the abnormal monitoring data is detected, determine the abnormal state information of the power transmission system based on the abnormal monitoring data.

[0071] In some examples, it is possible to determine whether each indicator is in an abnormal state by analyzing whether the various indicators in the anomaly monitoring data exceed the corresponding alarm threshold, thereby obtaining the corresponding anomaly status information.

[0072] In some examples, further anomaly analysis can be performed directly on the anomaly monitoring data to obtain the corresponding anomaly state information. For instance, a pre-trained anomaly state recognition model can be used to determine the anomaly state information based on the anomaly monitoring data. This anomaly state recognition model can be a model that has been trained to have the ability to predict anomaly state information by taking anomaly monitoring data as model input and anomaly state information as model output. During training, sample anomaly monitoring data can be used as sample data and input into the model, so that the model outputs predicted anomaly state information. Then, based on the difference between the predicted anomaly state information and the expected anomaly state represented by the state label, a general loss function is used to calculate the loss value. Then, based on the loss value, a general training algorithm (such as gradient descent) is used to train the model so that the trained model can have the above-mentioned capabilities.

[0073] In some examples, when the abnormal monitoring data is detected, the abnormal state information of the power transmission system can be determined based on the abnormal monitoring data.

[0074] In one alternative implementation, the at least one target is arranged in a preset order, the preset order being adapted to indicate the electrical signal transmission sequence between the at least one monitored component;

[0075] The step of obtaining real-time element information for each of the key state elements through target state analysis, based at least on each of the key state elements and each of the real-time monitoring data, includes:

[0076] Extract the corresponding real-time monitoring data features from each of the real-time monitoring data;

[0077] For each target, based at least on the real-time monitoring data features corresponding to it and the real-time monitoring data features corresponding to the targets preceding it, an artificial intelligence model is invoked to perform target state analysis, generate first element information corresponding to the target, and decompose the first element information to obtain real-time element information of each key state element corresponding to the target. The targets preceding it refer to the targets whose order in the preset order is earlier than the target.

[0078] In this embodiment, it is easy to understand that since the target preceding the target refers to the target whose order in the preset order is before the target, and the preset order is suitable for indicating the electrical signal transmission order, it is equivalent to combining the relevant real-time monitoring data features of the monitored component whose electrical signal transmission order is earlier with the generation process of the element information corresponding to the current monitored component. This allows the generated first element information to take the relevant real-time monitoring data of the monitored component whose electrical signal transmission order is earlier as a reference, combined with the actual position of the monitored component in the power transmission system, and considering the possible impacts that the electrical signal transmitted to the monitored component may have been affected in the past. This allows the first element information to more accurately reflect the actual situation of the electrical signal of the monitored component.

[0079] In some examples, the aforementioned real-time monitoring data features corresponding to the target and the real-time monitoring data features corresponding to the target preceding the target can be input into the aforementioned artificial intelligence model for target state analysis. Here, the artificial intelligence model can be pre-trained. This model can generate the first element information corresponding to the target based on the real-time monitoring data features corresponding to it and the real-time monitoring data features corresponding to the target preceding it. The artificial intelligence model can be a model that has been trained to have the predictive ability to use the real-time monitoring data features corresponding to it and the real-time monitoring data features corresponding to the target preceding it as model inputs and the first element information corresponding to the target as model outputs. In specific training, the monitoring data features corresponding to the first sample target and the monitoring data features corresponding to the second sample target (the second sample target is the target preceding the first sample target) can be used as sample data (the sample data also carries the expected corresponding element information label, which represents the corresponding expected element information). The predicted element information generated by the model based on the sample data is obtained. Based on the difference between the predicted element information and the expected element information, a general loss function is used to calculate the loss value. Based on the loss value, a general training algorithm (such as gradient descent) is used to train the model so that the trained model can have the above-mentioned ability.

[0080] In some examples, real-time monitoring data features can be extracted from the real-time monitoring data by encoding it. For example, this encoding process can be implemented using at least one of the following methods:

[0081] For numerical time-series data in real-time monitoring (such as continuously sampled sensor data like voltage, current, temperature, and pressure), this encoding process can be achieved by calculating statistics on data within a sliding window (such as data from the most recent preset time period). These statistics can include at least one of the following: basic statistical characteristics (such as mean, variance, maximum, minimum, and / or peak-to-peak values), frequency domain characteristics (such as the dominant frequency and / or energy spectral density after Fast Fourier Transform), and transient characteristics (such as zero-crossing rate, number of abrupt change points, and / or slope change rate). The corresponding real-time monitoring data features can include these statistics.

[0082] For waveform / oscillation signals in real-time monitoring data (such as harmonic signals in current data, vibration signals in vibration data, and / or partial discharge signals in current data, etc., which contain frequency components), the signals can be converted to the time-frequency plane and energy distribution features can be extracted through short-time Fourier transform and / or wavelet packet decomposition to achieve the encoding process (the corresponding real-time monitoring data features can contain the energy distribution features).

[0083] For the joint features of multiple related sensors (such as the joint features of voltage, current and temperature), a deep autoencoder neural network can be used to learn the low-dimensional implicit representation of the data from these multiple related sensors in order to achieve the encoding process (the corresponding real-time monitoring data features can contain this low-dimensional implicit representation).

[0084] In one optional implementation, the artificial intelligence model includes M feature interaction layers and M feature attention layers, where M is an integer greater than 1;

[0085] The process involves, based at least on the corresponding real-time monitoring data features and the real-time monitoring data features corresponding to targets preceding it, invoking an artificial intelligence model to perform target state analysis and generate the first element information corresponding to the target, including:

[0086] Based on the real-time monitoring data features corresponding to the target preceding the target, the first feature interaction layer is invoked to generate the first reference feature;

[0087] Based at least on the first reference feature and the corresponding real-time monitoring data feature, the first feature attention layer is invoked to generate the first attention feature;

[0088] Based at least on the m-th reference feature and its corresponding real-time monitoring data feature, the m-th feature attention layer is invoked to generate the m-th attention feature;

[0089] in,

[0090] m is an integer greater than 1 and not greater than M, and the m-th reference feature is generated based on the (m-1)-th attention feature by calling the m-th feature interaction layer;

[0091] The first element information corresponding to this target is generated based on the Mth attention feature.

[0092] It is understood that this embodiment is equivalent to alternating the connection of the M feature interaction layers and M feature attention layers included in the artificial intelligence model. That is, each pair of adjacent feature interaction layers is electrically connected via a feature attention layer, and no feature attention layer is connected before the first feature interaction layer, and a feature attention layer is electrically connected after the Mth feature interaction layer. In this way, the first feature interaction layer takes the real-time monitoring data feature corresponding to the target preceding the target as input, and the first feature attention layer takes the first reference feature and the real-time monitoring data feature corresponding to it as input. Then, the mth feature interaction layer takes the (m-1)th attention feature (i.e., the output of the (m-1)th feature attention layer) as input, and finally, the mth feature attention layer takes the mth reference feature (i.e., the output of the mth feature interaction layer) and the real-time monitoring data feature corresponding to it as input, thereby completing the construction of the artificial intelligence model.

[0093] In this embodiment, with the above settings, the interaction information output by the previous feature interaction layer can be obtained iteratively in each feature attention layer, so that the output of each feature attention layer can fully consider the supplementary information brought by the reference feature, thereby making the final m-th attention feature more accurate.

[0094] In some examples, the aforementioned feature interaction layer can be used to perform feature interaction processing on the various features in the input to obtain new composite features. This feature interaction processing can include at least one of the following: feature concatenation, feature vector multiplication. It is understood that since there may be one or more targets preceding the target, there may also be one or more real-time monitoring data features corresponding to the targets preceding the target. In this case, the first feature interaction layer can process these one or more features into the first reference feature. Subsequent feature interaction layers can then process the attention features in the input into reference features that conform to the input specifications of the feature attention layer.

[0095] In some examples, the aforementioned feature attention layer can be used to dynamically weight and fuse the various features in the input based on a cross-attention algorithm. This dynamic weighted fusion allows for deep feature fusion between the obtained reference features and attention features. This enables the AI ​​model to fully consider the feature meanings of other related real-time monitoring data features when processing real-time monitoring data features corresponding to the target, thereby improving the accuracy of the generated attention features corresponding to the target and further enhancing the accuracy of the generated first element information.

[0096] In some examples, the Mth attention feature can be decoded using a pre-defined decoder to generate the corresponding first element information. This decoder can be configured according to the encoding characteristics of the input features, which will not be elaborated here.

[0097] In one alternative implementation, for each of the stated objectives, the method further includes:

[0098] Determine the relative relationship between the monitored component corresponding to the target and the monitored component corresponding to each of the targets preceding the target, wherein the relative relationship is determined at least by the electrical signal transmission sequence between the two corresponding monitored components;

[0099] in,

[0100] The step of generating a first attention feature by calling the first feature attention layer based at least on the first reference feature and the corresponding real-time monitoring data feature includes: generating a first attention feature by calling the first feature attention layer based on the relative relationship, the first reference feature and the corresponding real-time monitoring data feature;

[0101] And / or,

[0102] The step of generating the mth attention feature by calling the mth feature attention layer based at least on the mth reference feature and the corresponding real-time monitoring data feature includes: generating the mth attention feature by calling the mth feature attention layer based on the relative relationship, the mth reference feature and the corresponding real-time monitoring data feature.

[0103] It is understood that the relative relationship determined by the sequence of electrical signal transmission may be taken into account in this embodiment. Therefore, the real-time monitoring data used at this time is generally data related to electrical signals, such as current and voltage.

[0104] In this embodiment, the feature attention layer can further acquire the above-mentioned relative relationships, so as to further consider the relative relationships (such as electrical signal transmission relationships) between the monitored components corresponding to different real-time monitoring data features when generating attention features, so as to improve the accuracy of the generated attention features corresponding to the target, and further improve the accuracy of the generated first element information.

[0105] In some examples, the relative relationship can be converted into a relative relationship feature that conforms to the input specification of the feature attention layer, and then input into the corresponding feature attention layer along with the corresponding reference feature and the real-time monitoring data feature mentioned above, so as to obtain the attention feature output by the corresponding feature attention layer.

[0106] In some examples, the above relative relationships may include the sequence of electrical signal transmission between the two monitored components.

[0107] In some examples, the aforementioned relative relationships can also be determined by the functional relationships (e.g., the sequential relationships between different functions), fault dependencies (e.g., whether a fault in a preceding monitored component will propagate to a subsequent monitored component), and / or state causality (e.g., whether a state change in a preceding monitored component will propagate to a subsequent monitored component). For example, the relative relationship may also include the aforementioned functional relationships, fault dependencies, and / or state causality.

[0108] In one optional implementation, the step of detecting whether there is abnormal monitoring data in the real-time monitoring data based on the difference between the element requirement information corresponding to each of the key state elements and the real-time element information includes:

[0109] For each of the key state elements, a first information feature of the corresponding element requirement information and a second information feature of the corresponding real-time element information are determined, and the similarity between the first information feature and the second information feature is determined, wherein the similarity is used to characterize the difference.

[0110] Based on the similarity, the system detects whether there is any abnormal monitoring data in the real-time monitoring data.

[0111] In some examples, the specific methods used to determine features in this embodiment can be found in other embodiments related to feature acquisition / extraction in this application, and will not be repeated here.

[0112] It should be noted that the similarity described in any one or more embodiments of this application can be calculated using at least one of the following: cosine similarity, Euclidean distance, Manhattan distance, Pearson correlation coefficient, etc. It should be understood that the implementation of this similarity method is merely illustrative and not intended to limit the scope of this application.

[0113] In some examples, the data corresponding to key state elements with similarity below a certain similarity threshold in the real-time monitoring data can be considered as anomaly monitoring data. In this case, if the similarity of each key state element is higher than the similarity threshold corresponding to that key state element, it can be considered that no anomaly monitoring data has been detected in the real-time monitoring data.

[0114] In one optional implementation, detecting the presence of abnormal monitoring data in the real-time monitoring data based on the similarity includes:

[0115] Based on the similarity, identify the elements to be identified from all the key state elements whose similarity is lower than the similarity threshold;

[0116] Extract the data of the elements to be identified that match the elements to be identified from the real-time monitoring data;

[0117] The system detects whether there is any abnormal monitoring data in the data of the elements to be identified.

[0118] In some examples, the elements to be identified can be understood as key state elements that have undergone initial screening. In this case, further refined anomaly detection can be performed on the data of these elements to accurately identify the presence of abnormal monitoring data through re-screening. This ensures detection accuracy while avoiding excessive anomaly detection on too much data. Here, it's easy to understand that more refined anomaly detection can refer to a larger number of indicators to be detected, or a more finely divided set of preset indicator values / ranges for comparison.

[0119] In one optional implementation, the data of the element to be identified refers to the real-time monitoring data of the target corresponding to the element to be identified.

[0120] In one alternative implementation, the at least one monitored component includes at least one first component of the trolley and / or at least one second component of the graphitization furnace.

[0121] At least one of the following first components of the power transmission vehicle includes at least one of the following: the output terminal of the power transmission vehicle, a conductive cable, and an electrode clamp;

[0122] At least one second component of the graphitization furnace includes at least one of the following: a graphite electrode of the graphitization furnace, or a conductive heating element inside the furnace.

[0123] In some examples, the monitored components include the output end of the trolley.

[0124] The real-time monitoring data used to characterize the target at the output terminal may include at least one of the following: the output voltage (750 V in this example under normal conditions), the output current (25 kA in this example under normal conditions), the current fluctuation rate (±3% in this example under normal conditions), and the coolant temperature (42°C in this example under normal conditions).

[0125] The key state elements used to characterize the target of the output terminal may include at least one of the following: power supply stability of the output terminal (e.g., characterized by the output voltage, output current and / or current fluctuation rate of the output terminal), output power accuracy (e.g., characterized by the output voltage, output current and / or current fluctuation rate of the output terminal), and heat dissipation capacity (e.g., characterized by the difference between the actual coolant temperature and the expected coolant temperature under the current operating conditions).

[0126] The key state elements required to characterize the target at the output terminal may include at least one of the following: output voltage deviation requirement (≤±2% under normal conditions in this example), current fluctuation rate requirement (≤±5% under normal conditions in this example), coolant temperature requirement (constantly <50°C under normal conditions in this example, i.e., less than 50°C regardless of the operating conditions), and continuous full-load operation time requirement (≥8 hours under normal conditions in this example).

[0127] In some examples, the real-time monitoring data for the conductive cable may include at least one of the following: surface temperature, insulation resistance, and voltage drop; the corresponding key status elements may include at least one of the following: temperature rise, insulation, and contact resistance; and the corresponding element requirement information may include at least one of the following: temperature requirement information and insulation requirement information.

[0128] In some examples, the real-time monitoring data corresponding to the electrode clamp may include at least one of the following: temperature, clamping force, and pressure drop; the corresponding key status elements may include at least one of the following: contact quality and clamping reliability; and the corresponding element requirement information may include at least one of the following: clamping force requirement information and pressure drop requirement information.

[0129] In some examples, the real-time monitoring data corresponding to the graphite electrode may include at least one of the following: current density, loss rate, and cracks; the corresponding key state elements may include at least one of the following: conductivity and structural integrity; and the corresponding element requirement information may include at least one of the following: loss requirement information (e.g., loss rate ≤ 0.15 mm / h) and crack requirement information (e.g., crack length ≤ 5 mm).

[0130] In some examples, the real-time monitoring data corresponding to the heating elements in the furnace may include at least one of the following: impedance, temperature distribution, and three-phase balance. The corresponding key state elements may include at least one of the following: temperature field uniformity and impedance stability. The corresponding element requirement information may include at least one of the following: impedance change rate requirement information (e.g., the impedance change rate relative to the initial value is required to be no more than ±10%), temperature difference requirement information (e.g., the temperature difference is required to be ≤ 500°C), and three-phase balance requirement information (e.g., the three-phase balance is required to be ≥95%).

[0131] In one optional embodiment, the at least one monitored component includes at least one first component of the power transmission vehicle and at least one second component of the graphitization furnace. The at least one first component of the power transmission vehicle includes the output end of the power transmission vehicle, a conductive cable, and an electrode clamp. The at least one second component of the graphitization furnace includes the graphite electrode of the graphitization furnace and a conductive heating element inside the furnace.

[0132] The electrical signal transmission sequence between at least one monitored component includes a preset electrical signal transmission sequence, and the order of each component corresponding to the preset electrical signal transmission sequence is as follows: the output terminal, the conductive cable, the electrode clamp, the graphite electrode, and the conductive heating element inside the furnace.

[0133] It is understood that in this embodiment, the output terminal, conductive cable, electrode clamp, graphite electrode, and conductive heating element inside the furnace are connected in sequence. The output terminal is used to convert the electrical energy stored in the power transmission vehicle and output it in sequence to the conductive cable, electrode clamp, graphite electrode, and conductive heating element inside the furnace, thereby realizing the sequential transmission of electrical signals.

[0134] Secondly, see Figure 2 The diagram shows a schematic of the power transmission system provided in an embodiment of this application. The power transmission system 200 includes a monitoring device 201, a graphitization furnace 202, and a power transmission vehicle 203. The monitoring device 201 is configured as follows:

[0135] During the process of the power transmission vehicle 203 supplying power to the graphitization furnace 202, real-time monitoring data of at least one target is acquired, wherein the at least one target is used to characterize at least one monitored component in the power transmission system.

[0136] Obtain the key state elements corresponding to each of the at least one target and the element requirement information of the key state elements;

[0137] Based at least on each of the key state elements and each of the real-time monitoring data, real-time element information of each of the key state elements is obtained through target state analysis;

[0138] Based on the differences between the element requirement information and the real-time element information corresponding to each of the key state elements, the presence of abnormal monitoring data is detected in the real-time monitoring data.

[0139] If the abnormal monitoring data is detected, the abnormal status information of the power transmission system is determined based on the abnormal monitoring data.

[0140] In one alternative implementation, the at least one target is arranged in a preset order, the preset order being adapted to indicate the electrical signal transmission sequence between the at least one monitored component;

[0141] The step of obtaining real-time element information for each of the key state elements through target state analysis, based at least on each of the key state elements and each of the real-time monitoring data, includes:

[0142] Extract the corresponding real-time monitoring data features from each of the real-time monitoring data;

[0143] For each target, based at least on the real-time monitoring data features corresponding to it and the real-time monitoring data features corresponding to the targets preceding it, an artificial intelligence model is invoked to perform target state analysis, generate first element information corresponding to the target, and decompose the first element information to obtain real-time element information of each key state element corresponding to the target. The targets preceding it refer to the targets whose order in the preset order is earlier than the target.

[0144] In one optional implementation, the artificial intelligence model includes M feature interaction layers and M feature attention layers, where M is an integer greater than 1;

[0145] The process involves, based at least on the corresponding real-time monitoring data features and the real-time monitoring data features corresponding to targets preceding it, invoking an artificial intelligence model to perform target state analysis and generate the first element information corresponding to the target, including:

[0146] Based on the real-time monitoring data features corresponding to the target preceding the target, the first feature interaction layer is invoked to generate the first reference feature;

[0147] Based at least on the first reference feature and the corresponding real-time monitoring data feature, the first feature attention layer is invoked to generate the first attention feature;

[0148] Based at least on the m-th reference feature and its corresponding real-time monitoring data feature, the m-th feature attention layer is invoked to generate the m-th attention feature;

[0149] in,

[0150] m is an integer greater than 1 and not greater than M, and the m-th reference feature is generated based on the (m-1)-th attention feature by calling the m-th feature interaction layer;

[0151] The first element information corresponding to this target is generated based on the Mth attention feature.

[0152] In an alternative implementation, for each of the targets, the monitoring device 201 is further configured to:

[0153] Determine the relative relationship between the monitored component corresponding to the target and the monitored component corresponding to each of the targets preceding the target, wherein the relative relationship is determined at least by the electrical signal transmission sequence between the two corresponding monitored components;

[0154] in,

[0155] The step of generating a first attention feature by calling the first feature attention layer based at least on the first reference feature and the corresponding real-time monitoring data feature includes: generating a first attention feature by calling the first feature attention layer based on the relative relationship, the first reference feature and the corresponding real-time monitoring data feature;

[0156] And / or,

[0157] The step of generating the mth attention feature by calling the mth feature attention layer based at least on the mth reference feature and the corresponding real-time monitoring data feature includes: generating the mth attention feature by calling the mth feature attention layer based on the relative relationship, the mth reference feature and the corresponding real-time monitoring data feature.

[0158] In one optional implementation, the step of detecting whether there is abnormal monitoring data in the real-time monitoring data based on the difference between the element requirement information corresponding to each of the key state elements and the real-time element information includes:

[0159] For each of the key state elements, a first information feature of the corresponding element requirement information and a second information feature of the corresponding real-time element information are determined, and the similarity between the first information feature and the second information feature is determined, wherein the similarity is used to characterize the difference.

[0160] Based on the similarity, the system detects whether there is any abnormal monitoring data in the real-time monitoring data.

[0161] In one optional implementation, detecting the presence of abnormal monitoring data in the real-time monitoring data based on the similarity includes:

[0162] Based on the similarity, identify the elements to be identified from all the key state elements whose similarity is lower than the similarity threshold;

[0163] Extract the data of the elements to be identified that match the elements to be identified from the real-time monitoring data;

[0164] The system detects whether there is any abnormal monitoring data in the data of the elements to be identified.

[0165] In one optional implementation, the data of the element to be identified refers to the real-time monitoring data of the target corresponding to the element to be identified.

[0166] In one optional implementation, the at least one monitored component includes at least one first component of the power transmission vehicle 203 and / or at least one second component of the graphitization furnace 202;

[0167] At least one of the first components of the trolley 203 includes at least one of the following: the output terminal of the trolley 203, a conductive cable, and an electrode clamp;

[0168] At least one second component of the graphitization furnace 202 includes at least one of the following: a graphite electrode of the graphitization furnace 202, or a conductive heating element inside the furnace.

[0169] In one optional embodiment, the at least one monitored component includes at least one first component of the power transmission vehicle 203 and at least one second component of the graphitization furnace 202. The at least one first component of the power transmission vehicle 203 includes the output terminal of the power transmission vehicle 203, a conductive cable, and an electrode clamp. The at least one second component of the graphitization furnace 202 includes the graphite electrode of the graphitization furnace 202 and a conductive heating element inside the furnace.

[0170] The electrical signal transmission sequence between at least one monitored component includes a preset electrical signal transmission sequence, and the order of each component corresponding to the preset electrical signal transmission sequence is as follows: the output terminal, the conductive cable, the electrode clamp, the graphite electrode, and the conductive heating element inside the furnace.

[0171] Thirdly, embodiments of this application provide a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described in any of the above-mentioned embodiments.

[0172] Fourthly, embodiments of this application provide a computer program product, including computer instructions that, when executed by a processor, implement the steps of the method described in any of the above-described embodiments.

[0173] Fifthly, embodiments of this application provide a computer device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the steps of the method described in any of the preceding claims.

[0174] See Figure 3 The computer device in this embodiment includes a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301, such as a power transmission system status monitoring program. When the processor 301 executes the computer program, it implements the steps in the various power transmission system status monitoring method embodiments described above.

[0175] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 302 and executed by the processor 301 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the computer device.

[0176] The computer device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will understand that the schematic diagram is merely an example of a computer device and does not constitute a limitation on the computer device. It may include more or fewer components than shown, or combine certain components, or different components. For example, the computer device may also include input / output devices, network access devices, buses, etc.

[0177] The processor 301 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor 301 can be any conventional processor. The processor 301 is the control center of the computer device, connecting various parts of the entire computer device through various interfaces and lines.

[0178] The memory 302 can be used to store the computer programs and / or modules. The processor 301 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 302 and calling the data stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 302 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0179] Wherein, if the modules / units integrated into the computer device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a non-transitory computer-readable storage medium. When the computer program is executed by the processor 301, it can implement the steps of the various method embodiments described above. Wherein, the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0180] In summary, the embodiments of this application have at least the following beneficial effects:

[0181] In this embodiment of the application, during the process of the power transmission vehicle supplying power to the graphitization furnace, real-time monitoring data of at least one target is acquired, wherein the at least one target is used to characterize at least one monitored component in the power transmission system; key state elements corresponding to each of the at least one target and element requirement information of the key state elements are acquired; based at least on each key state element and each of the real-time monitoring data, real-time element information of each key state element is obtained through target state analysis; based on the differences between the element requirement information and the real-time element information corresponding to each of the key state elements, abnormal monitoring data is detected in the real-time monitoring data; if abnormal monitoring data is detected, abnormal state information of the power transmission system is determined based on the abnormal monitoring data. This allows for the initial screening of real-time monitoring data by first obtaining real-time element information of different key state elements through target state analysis, then comparing the differences between the real-time element information and the corresponding element requirement information, and finally analyzing the detected abnormal monitoring data to obtain the corresponding abnormal state information. This enables accurate identification of abnormal states in the initially screened data, thereby improving equipment safety by accurately monitoring the abnormal states of the power transmission system.

[0182] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary hardware platforms, or it can be implemented entirely by hardware. Based on this understanding, all or part of the technical solutions of this application that contribute to the background technology can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM (Read-Only Memory) / RAM (Random Access Memory), magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0183] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications are also considered to be within the scope of protection of this application.

Claims

1. A method of monitoring a state of a power transmission system, characterized by, The power transmission system comprises a graphitization furnace and a power transmission vehicle, and the method comprises: In the process of the power transmission vehicle transmitting power to the graphitization furnace, real-time monitoring data of at least one target is obtained, wherein the at least one target is used to represent at least one monitored component in the power transmission system; Corresponding key state elements of the at least one target and element requirement information of the key state elements are obtained; Real-time element information of each key state element is obtained through target state analysis based on the key state elements and the real-time monitoring data; Based on the differences between the element requirement information and the real-time element information of all the key state elements, it is detected whether there is abnormal monitoring data in the real-time monitoring data; In the case of detecting the abnormal monitoring data, the abnormal state information of the power transmission system is determined based on the abnormal monitoring data; The at least one target is arranged in a preset order, and the preset order is suitable for indicating the electrical signal transmission sequence between the at least one monitored component; The real-time element information of each key state element is obtained through target state analysis based on the key state elements and the real-time monitoring data, comprising: Real-time monitoring data features corresponding to each real-time monitoring data are extracted; For each target, an artificial intelligence model is called for target state analysis based on the real-time monitoring data features corresponding to the target and the real-time monitoring data features corresponding to the target before the target, to generate first element information corresponding to the target, and the first element information is split to obtain real-time element information of each key state element corresponding to the target, wherein the target before the target refers to the target before the target in the preset order.

2. The method of claim 1, wherein, The artificial intelligence model comprises M feature interaction layers and M feature attention layers, and M is an integer greater than 1; The artificial intelligence model is called for target state analysis based on the real-time monitoring data features corresponding to the target and the real-time monitoring data features corresponding to the target before the target to generate first element information corresponding to the target, comprising: Based on the real-time monitoring data features corresponding to the target before the target, a first reference feature is generated by calling a first feature interaction layer; Based on the first reference feature and the real-time monitoring data features corresponding to the target, a first attention feature is generated by calling a first feature attention layer; Based on the mth reference feature and the real-time monitoring data features corresponding to the target, an mth attention feature is generated by calling an mth feature attention layer; Wherein, m is an integer greater than 1 and not greater than M, and the mth reference feature is generated by calling an mth feature interaction layer based on the (m-1)th attention feature; The first element information corresponding to the target is generated according to the Mth attention feature.

3. The method of claim 2, wherein, For each target, the method further comprises: determining a relative relationship between the target and each of the targets before the target, wherein the relative relationship is determined by at least an order of electrical signal transmission between the corresponding two monitored components; wherein, the calling of the first feature attention layer to generate the first attention feature based on the first reference feature and the real-time monitoring data feature corresponding thereto comprises: calling the first feature attention layer to generate the first attention feature based on the relative relationship, the first reference feature and the real-time monitoring data feature corresponding thereto; and / or, the calling of the mth feature attention layer to generate the mth attention feature based on the mth reference feature and the real-time monitoring data feature corresponding thereto comprises: calling the mth feature attention layer to generate the mth attention feature based on the relative relationship, the mth reference feature and the real-time monitoring data feature corresponding thereto.

4. The method of claim 1, wherein, the detecting of whether there is abnormal monitoring data in the real-time monitoring data based on the difference between the element requirement information corresponding to each of the key state elements and the real-time element information comprises: for each of the key state elements, determining a first information feature of the corresponding element requirement information and a second information feature of the corresponding real-time element information, and determining a similarity between the first information feature and the second information feature, wherein the similarity is used to represent the difference; the detecting of whether there is abnormal monitoring data in the real-time monitoring data based on the similarity comprises:

5. The method of claim 4, wherein, based on the similarity, determining a to-be-screened element whose similarity is lower than a similarity threshold from all the key state elements; extracting to-be-screened element data matching the to-be-screened element from the real-time monitoring data; detecting whether there is abnormal monitoring data in the to-be-screened element data.

6. The method of claim 5, wherein the to-be-screened element data refers to the real-time monitoring data of the target corresponding to the to-be-screened element.

7. The method of any one of claims 1-6, wherein the at least one monitored component comprises at least one first component of the power supply vehicle and / or at least one second component of the graphitization furnace; the at least one first component of the power supply vehicle comprises at least one of: an output end of the power supply vehicle, a conductive cable, and an electrode clamp; the at least one second component of the graphitization furnace comprises at least one of: a graphite electrode of the graphitization furnace and an in-furnace conductive heating element.

8. The method of claim 7, wherein ​ ​ ​ The at least one monitored component includes at least one first component of the power feeding vehicle and at least one second component of the graphitization furnace, the at least one first component of the power feeding vehicle includes an output end, a conductive cable and an electrode clamp of the power feeding vehicle, and the at least one second component of the graphitization furnace includes a graphite electrode and an in-furnace conductive heating element of the graphitization furnace. The electrical signal transmission sequence between the at least one monitored component includes a preset electrical signal transmission sequence, and the sequence of each component corresponding to the preset electrical signal transmission sequence is in turn the output end, the conductive cable, the electrode clamp, the graphite electrode and the in-furnace conductive heating element.

9. A power transmission system, characterized by comprising: The power feeding system includes a monitoring device, a graphitization furnace and a power feeding vehicle, and the monitoring device is configured to: During the process of feeding power from the power feeding vehicle to the graphitization furnace, real-time monitoring data of at least one target is obtained, wherein the at least one target is used to represent at least one monitored component in the power feeding system; Corresponding key state elements and element requirement information of the key state elements of the at least one target are obtained; Real-time element information of each key state element is obtained through target state analysis based on the key state elements and the real-time monitoring data; Based on the differences between the element requirement information and the real-time element information of all the key state elements, it is detected whether there is abnormal monitoring data in the real-time monitoring data; In the case of detecting the abnormal monitoring data, the abnormal state information of the power feeding system is determined based on the abnormal monitoring data; The at least one target is arranged in a preset sequence, and the preset sequence is suitable for indicating the electrical signal transmission sequence between the at least one monitored component. The real-time element information of each key state element is obtained through target state analysis based on the key state elements and the real-time monitoring data, including: Real-time monitoring data features corresponding to each real-time monitoring data are extracted from the real-time monitoring data; For each target, an artificial intelligence model is called for target state analysis based on the real-time monitoring data features corresponding to the target and the real-time monitoring data features corresponding to the target before the target, to generate first element information corresponding to the target, and the first element information is split to obtain real-time element information of each key state element corresponding to the target, wherein the target before the target refers to a target whose sequence in the preset sequence is before the target.

Images