Substation dynamic adjustment method and device, electronic equipment and computer readable medium

Abstract

Embodiments of the present disclosure disclose a power substation dynamic adjustment method and device, electronic equipment and a computer readable medium. A specific embodiment of the method comprises: obtaining a set of multi-element operating state information, a set of multi-element collaborative construction information and a set of multi-element target value information of a target power substation and performing de-privacy processing to obtain a set of de-privacy power substation multi-element information; determining a set of associated prediction index information; determining a set of index entropy weight values and a set of index risk weight values; generating a set of index target weight values; performing state prediction processing on the target power substation to obtain power substation state prediction information; performing operating state attribution analysis on the target power substation to obtain power substation state attribution information; and dynamically adjusting the target power substation according to the power substation state attribution information. The embodiment can improve the evaluation accuracy of the multi-element power substation, reduce the loss and cost of the multi-element power substation and improve the stability of the target power substation by constructing comprehensive and diversified evaluation indexes of the power substation.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of computer technology, and in particular, to a substation dynamic adjustment method and device, electronic equipment and a computer readable medium. BACKGROUND

[0002] With the accelerated construction of new power systems, the collaborative construction of diversified substations (such as the collaborative integration of multiple types of facilities such as new energy power stations, energy storage stations, and smart substations) has become a core link to improve the flexibility, reliability, and economy of the power system. Scientific and accurate evaluation of diversified substations is an important basis for guiding the planning, construction, and operation of diversified substations. For dynamic adjustment of a diversified substation, a commonly used method is to obtain a set of operation state information and a set of target value attribute information of a new power system diversified substation. Then, the state weight and the target value weight of the set of operation state information and the set of target value attribute information are determined using expert experience. Finally, the state weight and the target value weight are weighted and summed to obtain system evaluation information of the diversified substation, and the diversified substation is dynamically adjusted according to the system evaluation information.

[0003] However, in practice, it is found that when the above method is used to dynamically adjust the diversified substation, the following technical problems often exist: Since the expert experience has a certain subjectivity, and the expert experience cannot be updated in a timely manner over time, there is a certain lag, and the diversified substation is evaluated from two aspects of the set of operation state information and the set of target value attribute information, focusing on a single performance evaluation, there is a certain one-sidedness, resulting in low accuracy of the diversified substation construction and operation evaluation information, which cannot accurately guide and adjust the diversified substation, causing low stability of the diversified substation, and low damage rate of the diversified substation.

[0004] The above information disclosed in this BACKGROUND section is only for the purpose of enhancing the understanding of the background of the present disclosure and, therefore, can include information that does not form the prior art known to those of ordinary skill in the art in the country. SUMMARY

[0005] The summary of the present disclosure is intended to introduce the concepts in a simplified form, which will be described in detail in the following detailed description. The summary of the present disclosure is not intended to identify key or essential features of the claimed technology nor is it intended to be used to limit the scope of the claimed technology.

[0006] Some embodiments of the present disclosure propose a substation dynamic adjustment method and device, electronic equipment and a computer readable medium to solve one or more of the technical problems mentioned in the above BACKGROUND section.

[0007] In a first aspect, some embodiments of this disclosure provide a method for dynamic adjustment of a substation, comprising: acquiring a multi-dimensional operating status information set, a multi-dimensional collaborative construction information set, and a multi-dimensional target value information set of a target substation; performing privacy-removing processing on the multi-dimensional operating status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set to obtain a privacy-removed substation multi-dimensional information set; determining a set of related prediction indicators for the target substation based on the privacy-removed substation multi-dimensional information set; determining an indicator entropy weight value set and an indicator risk weight value set of the related prediction indicator information set based on the privacy-removed substation multi-dimensional information set; generating an indicator target weight value set for the related prediction indicator information set based on the indicator entropy weight value set and the indicator risk weight value set; performing state prediction processing on the target substation based on the indicator target weight value set and the related prediction indicator information set to obtain substation state prediction information; in response to determining that the substation state prediction information does not meet preset state prediction conditions, performing an operating state attribution analysis on the target substation to obtain substation state attribution information; and dynamically adjusting the target substation based on the substation state attribution information.

[0008] Secondly, some embodiments of this disclosure provide a substation dynamic adjustment device, comprising: an acquisition unit configured to acquire a multi-dimensional operating status information set, a multi-dimensional collaborative construction information set, and a multi-dimensional target value information set of a target substation; a privacy removal unit configured to perform privacy removal processing on the multi-dimensional operating status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set to obtain a privacy-removed substation multi-dimensional information set; a first determination unit configured to determine a set of associated prediction index information for the target substation based on the privacy-removed substation multi-dimensional information set; and a second determination unit configured to determine the index entropy weight of the associated prediction index information set based on the privacy-removed substation multi-dimensional information set. The system includes: a value set and an indicator risk weight value set; a generation unit configured to generate an indicator target weight value set for the associated prediction indicator information set based on the indicator entropy weight value set and the indicator risk weight value set; a state prediction unit configured to perform state prediction processing on the target substation based on the indicator target weight value set and the associated prediction indicator information set to obtain substation state prediction information; an attribution analysis unit configured to perform operational state attribution analysis on the target substation in response to determining that the substation state prediction information does not meet preset state prediction conditions to obtain substation state attribution information; and a dynamic adjustment unit configured to dynamically adjust the target substation based on the substation state attribution information.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0011] The various embodiments of this disclosure have the following beneficial effects: The substation dynamic adjustment method of some embodiments of this disclosure can improve the accuracy of multi-element substation evaluation, reduce the loss and cost of multi-element substations, and improve the stability of the target substation. Specifically, the reason why the accuracy of relevant multi-element substation construction and operation evaluation information is low, and why it cannot provide precise guidance and adjustment for multi-element substations, resulting in low stability and low damage rate of multi-element substations, is that: the use of expert experience has a certain degree of subjectivity, and expert experience cannot be updated in a timely manner, resulting in a certain lag; moreover, evaluating multi-element substations from two aspects, namely the operating status information set and the target value attribute information set, focuses on a single performance evaluation, which has a certain degree of one-sidedness. Based on this, the substation dynamic adjustment method of some embodiments of this disclosure can first obtain the multi-element operating status information set, the multi-element collaborative construction information set, and the multi-element target value information set of the target substation. Here, obtaining multi-faceted data on the operating status, collaborative construction, and target value of the target substation facilitates subsequent accurate evaluation of the target substation. Secondly, the aforementioned multi-dimensional operational status information set, multi-dimensional collaborative construction information set, and multi-dimensional target value information set are subjected to privacy-de-identification processing to obtain a privacy-de-identified substation multi-dimensional information set. Here, privacy-de-identification processing improves the data security of the target substation and reduces the leakage of private data. Thirdly, based on the aforementioned privacy-de-identified substation multi-dimensional information set, a set of related prediction indicators for the target substation is determined. Here, the privacy-de-identified substation multi-dimensional information set is transformed into an interpretable and quantifiable set of indicators, facilitating the subsequent determination of the weight values ​​of the indicator set. Next, based on the aforementioned privacy-de-identified substation multi-dimensional information set, the set of indicator entropy weight values ​​and indicator risk weight values ​​for the aforementioned related prediction indicator information set are determined. Here, the related prediction indicator information set is quantified using both entropy and risk values, which complement each other, avoiding the limitations of the first weight and improving the comprehensiveness and accuracy of indicator quantification. Subsequently, based on the aforementioned indicator entropy weight value set and indicator risk weight value set, a set of target weight values ​​for the related prediction indicator information set is generated. Here, both entropy and risk values ​​can be considered, avoiding the bias caused by the first weight, resulting in a comprehensive and balanced target weight value, which facilitates subsequent state prediction processing of the target substation. Then, based on the aforementioned target weight value set and the aforementioned associated prediction indicator information set, state prediction processing is performed on the target substation to obtain substation state prediction information. Here, transforming the multi-scale associated prediction indicator information set and the multi-dimensional target weight value set into more intuitive representations of the target substation's state information enables accurate comprehensive evaluation of the target substation, making the substation state prediction information more consistent with the actual situation of the target substation. Then, in response to the determination that the aforementioned substation state prediction information does not meet the preset state prediction conditions, an operational state attribution analysis is performed on the target substation to obtain substation state attribution information.Here, when the target substation fails to meet preset conditions, root cause analysis is performed to trace the more specific reasons for the non-compliance, providing precise aspects for subsequent adjustments to the target substation. Finally, based on the aforementioned substation status attribution information, the target substation is dynamically adjusted. Here, precise adjustments using the provided adjustment aspects can improve the stability of the target substation and reduce its failure rate. Therefore, this dynamic substation adjustment method can improve the accuracy of multi-substation assessments and provide precise adjustment aspects when preset conditions are not met, thereby reducing losses and adjustment costs for multi-substations and improving the stability of the target substation. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a flowchart of some embodiments of the substation dynamic adjustment method according to this disclosure;

[0014] Figure 2 This is a structural schematic diagram of some embodiments of the substation dynamic adjustment device according to the present disclosure;

[0015] Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0016] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0017] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0018] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0019] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0020] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0021] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] Figure 1 A flowchart 100 is shown, illustrating some embodiments of a substation dynamic adjustment method according to the present disclosure. The substation dynamic adjustment method includes the following steps:

[0023] Step 101: Obtain the multi-dimensional operating status information set, multi-dimensional collaborative construction information set, and multi-dimensional target value information set of the target substation.

[0024] In some embodiments, the executing entity (e.g., electronic equipment) of the above-mentioned substation dynamic adjustment method can acquire a multi-dimensional operating status information set, a multi-dimensional collaborative construction information set, and a multi-dimensional target value information set of the substation through wired or wireless connections. The target substation can be a substation collaboratively constructed by integrating various intelligent devices such as new energy power plants and energy storage stations. The multi-dimensional operating status information in the above-mentioned multi-dimensional operating status information set can be information characterizing the overall operating status of the target substation system. For example, the multi-dimensional operating status information set can include, but is not limited to, at least one of the following: substation equipment reliability, substation power quality, substation operating efficiency, and substation failure probability. The multi-dimensional collaborative construction information in the above-mentioned multi-dimensional collaborative construction information set can be information characterizing the collaborative interaction, cooperation, and support between the target substation and external power systems. The external power system can be a distribution network, energy storage system, load system, or distributed power generation system. For example, the multi-dimensional collaborative construction information can include, but is not limited to, at least one of the following: power reserve sharing rate, data communication information, multi-system collaborative control response speed, power mutual assistance capability information, and voltage support information. The multi-objective value information in the aforementioned multi-objective value information set can be comprehensive value information used to assess the economic, social, and environmental impacts of the target substation. For example, the aforementioned multi-objective value information set may include, but is not limited to, at least one of the following: power supply reliability value information, renewable energy absorption capacity information, carbon emission information, land resource utilization rate information, network loss value information (network loss benefits), infrastructure value information (substation investment and construction costs), and acquisition value information (investment return benefits).

[0025] Step 102 involves performing privacy-de-identifying processing on the multi-dimensional operational status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set to obtain a privacy-de-identifying multi-dimensional information set for the substation.

[0026] In some embodiments, the aforementioned implementing entity may perform privacy-de-identifying processing on the aforementioned multi-dimensional operational status information set, the aforementioned multi-dimensional collaborative construction information set, and the aforementioned multi-dimensional target value information set to obtain a privacy-de-identifying substation multi-dimensional information set. The privacy-de-identifying substation multi-dimensional information in the aforementioned privacy-de-identifying substation multi-dimensional information set may be multimodal information on operation, system construction, and value aspects that does not contain privacy data and is used to evaluate the overall benefits of the power grid.

[0027] In some optional implementations of certain embodiments, the above-mentioned de-privacy processing of the above-mentioned multi-dimensional operating status information set, the above-mentioned multi-dimensional collaborative construction information set, and the above-mentioned multi-dimensional target value information set to obtain a privacy-de-privacy substation multi-dimensional information set may include the following steps:

[0028] The first step involves filtering out unstructured information from the aforementioned multi-dimensional operational status information set, multi-dimensional collaborative construction information set, and multi-dimensional target value information set to obtain a multi-dimensional substation unstructured information set. Specifically, the multi-dimensional substation unstructured information in this set can be information of unstructured data type from the aforementioned multi-dimensional operational status information set, multi-dimensional collaborative construction information set, and multi-dimensional target value information set.

[0029] The second step involves performing named entity recognition on the aforementioned multi-source substation unstructured information set based on a pre-defined substation knowledge graph, resulting in a substation entity information set. The substation entity information in this set can be instances with explicit names from the multi-source substation unstructured information set. For example, the substation entity information set may include, but is not limited to, at least one of the following: person's name, address, meter number, equipment information, and load information. The pre-defined substation knowledge graph can be an existing knowledge graph related to the substation field.

[0030] As an example, the aforementioned execution entity can first use a named entity recognition model to perform entity recognition on the aforementioned unstructured information set of the multi-substation, obtaining an initial entity information set. The named entity recognition model can be a deep neural network model that performs entity recognition on the input unstructured information set of the multi-substation. This named entity recognition model can be a deep neural network composed of a BERT (Bidirectional Encoder Representations from Transformers, a pre-trained language model based on the Transformer architecture), a GRU (Gate Recurrent Unit), and a CRF (Conditional Random Fields) model concatenated. Then, by using a pre-defined substation knowledge graph, the initial entity information set is resolved through referencing, yielding the substation entity information set.

[0031] The third step involves extracting contextual semantic features from the aforementioned substation entity information set to obtain a substation semantic feature vector set. The substation semantic feature vectors in this set represent the semantic information of the substation entity information within its context. In practice, the executing entity can first utilize a T5 (Text-to-Text Transfer Transformer) model to extract semantic features from the substation entity information set, obtaining a semantic feature vector set. Secondly, dependency parsing is performed on the aforementioned multi-dimensional unstructured substation information set to obtain a multi-dimensional parse tree. This multi-dimensional parse tree can represent the syntactic structure of the multi-dimensional unstructured substation information set in tree form. Nodes in the multi-dimensional parse tree can represent substation entity information, and edges can represent dependencies between these entities. Finally, a graph attention network is used to encode the semantic feature vector set and the multi-dimensional parse tree to obtain the substation semantic feature vector set.

[0032] The fourth step is to determine the entity semantic annotation information set for each substation entity in the aforementioned substation entity information set. This entity semantic annotation information set can be information about the semantic roles of the substation entity information. In practice, the executing entity can utilize a semantic role labeling algorithm to determine the entity semantic annotation information set for each substation entity in the aforementioned substation entity information set. It should be noted that determining the entity semantic annotation information set by analyzing the semantic roles of substation entity information in the multi-dimensional substation unstructured information set can identify implicit relationships between entities. For example, if Zhang San (user) exceeds the electricity load limit (electricity consumption behavior) in XX community (address), role classification can implicitly associate the user with sensitive information such as address and behavior.

[0033] The fifth step involves performing feature embedding processing on the aforementioned entity semantic annotation information set and the aforementioned substation semantic feature vector set to obtain a substation text embedding feature vector set. The substation text embedding feature vectors in this set can represent the semantic information of the aforementioned substation entity information in different statements and the hidden entity relationships. This feature embedding processing can be performed through an embedding layer.

[0034] The sixth step involves inputting the aforementioned substation text embedding feature vector set into a privacy recognition model to obtain unstructured privacy data. This privacy recognition model can be a deep neural network that identifies the privacy data within the input substation text embedding feature vector set. This privacy recognition model can be a model composed of a Transformer model, a multi-head self-attention mechanism network, and a fully connected layer containing a Sigmoid function, connected in series.

[0035] The seventh step involves de-privacy processing of the aforementioned unstructured privacy data to obtain a de-privacy substation multivariate information set. This de-privacy processing can be performed through substitution or anonymization.

[0036] Optionally, the above-mentioned privacy-de-identifying processing of the aforementioned multi-dimensional operational status information set, the aforementioned multi-dimensional collaborative construction information set, and the aforementioned multi-dimensional target value information set to obtain a privacy-de-identified substation multi-dimensional information set may include the following steps:

[0037] The first step is to remove the unstructured information set of the substation from the aforementioned multi-dimensional operational status information set, multi-dimensional collaborative construction information set, and multi-dimensional target value information set to obtain the substation structured information set. The structured information in this substation structured information set can be information of structured data type from the aforementioned multi-dimensional operational status information set, multi-dimensional collaborative construction information set, and multi-dimensional target value information set.

[0038] The second step is to determine the structured information entropy set and the structured discrete entropy set of the aforementioned substation structured information set. The structured information entropy in the structured information entropy set can be a numerical value used to determine the uncertainty of the substation's structured information. The structured discrete entropy in the structured discrete entropy set can characterize the state where information uncertainty or mixed states are maximized. In practice, the executing entity can use the information entropy calculation formula and the maximum discrete entropy theorem to determine the structured information entropy set and the structured discrete entropy set of the aforementioned substation structured information set.

[0039] The third step involves generating a structured privacy set for the substation structured information set based on the aforementioned structured information entropy set and structured discrete entropy set. The structured privacy value characterizes the probability that the substation structured information is private data. It can also be an indicator of the degree of disorder in the substation structured information set. A higher structured privacy value indicates a greater probability that the substation structured information is private data. For example, the executing entity can first determine the difference between each structured information entropy in the structured information entropy set and the corresponding structured discrete entropy in the structured discrete entropy set, obtaining a structured entropy difference set. Then, the ratio of each structured entropy difference in the structured entropy difference set to the corresponding structured discrete entropy in the structured discrete entropy set is used as the structured privacy value, resulting in the structured privacy set.

[0040] The fourth step involves performing privacy-themed clustering on the substation structured information set based on the aforementioned structured privacy level set, resulting in a set of structured privacy data clusters and a set of structured non-privacy data clusters. Specifically, the structured privacy data clusters in the structured privacy data cluster set can be clusters composed of substation structured information corresponding to at least one structured privacy level greater than or equal to a preset privacy threshold. The structured non-privacy data clusters in the structured non-privacy data cluster set can be clusters composed of at least one substation structured information with a privacy level less than a preset privacy threshold. The preset privacy threshold can be a pre-defined adjacency value used to determine whether data is privacy-sensitive. For example, the preset privacy threshold could be 0.75. As an example, the executing entity can utilize a density-based clustering algorithm to perform privacy-themed clustering on the substation structured information set based on the aforementioned structured privacy level set, resulting in a set of structured privacy data clusters and a set of structured non-privacy data clusters.

[0041] The fifth step is to determine the cluster association information set between the aforementioned structured privacy data cluster set and the aforementioned structured non-privacy data cluster set. The cluster association information set can be inferred from the structured non-privacy data clusters to determine the association information set between the structured privacy data cluster sets. In practice, the implementing entity can utilize FP-Growth (Frequent Pattern Growth) frequent pattern trees to mine multidimensional relationships between clusters, thereby determining the cluster association information set between the aforementioned structured privacy data cluster set and the aforementioned structured non-privacy data cluster set.

[0042] Step 6: Based on the above cluster association information set, perform cluster update processing on the above structured privacy data cluster set and the above structured non-privacy data cluster set to obtain the updated structured privacy data cluster set.

[0043] As an example, the aforementioned execution entity can first filter at least one cluster association information from the aforementioned cluster association information set that is greater than or equal to a preset association threshold. The preset association threshold can be a pre-set critical value for user-implemented filtering. For example, the preset association threshold can be 0.65. Secondly, in response to determining that the ratio of the number of cluster association information included in the aforementioned at least one cluster association information to the total number is greater than or equal to a preset ratio threshold, for each substation structured information in the aforementioned substation structured information set, the following average value determination steps are performed: First, determine the substation structured information that has data association information with the aforementioned substation structured information, as associated structured information, to obtain an associated structured information set. The total number can be the number of cluster association information included in the aforementioned cluster association information set. The preset ratio threshold can be a pre-set value used for cluster updates. For example, the preset ratio threshold can be 0.7. Second, determine the conditional information entropy of the aforementioned substation structured information under the association conditions of each associated structured information in the aforementioned associated structured information set, to obtain a conditional information entropy set. The third step involves determining the ratio of the difference between the structured discrete entropy of the substation's structured information and the conditional information entropy set, to the structured discrete entropy, as the updated structured privacy level, thus obtaining the updated structured privacy level set. The fourth step is to determine the average value of the updated structured privacy level set, obtaining the structured privacy mean. Finally, the obtained structured privacy mean set is clustered to obtain the updated structured privacy data cluster set.

[0044] Step 7: Perform privacy-de-identification processing on the updated structured privacy data clusters to obtain a privacy-de-identified substation multi-source information set. This privacy-de-identification processing can be performed through substitution or anonymization.

[0045] Step 103: Based on the multivariate information set of privacy-removing substations, determine the associated prediction indicator information set for the target substation.

[0046] In some embodiments, the aforementioned executing entity may determine a set of associated predictive indicators for the target substation based on the aforementioned privacy-de-identified substation multivariate information set. The associated predictive indicators in this set may be statistical measures used to quantify the target substation's performance in terms of substation operation, collaboration with external systems, and value gains, reflecting the overall situation of the target substation. This set of associated predictive indicators may include, but is not limited to, at least one of the following: substation equipment reliability indicators, substation fault reliability indicators, substation automation level indicators, power resource dispatch indicators, information interaction delay indicators, acquisition value indicators, and operation and maintenance value indicators.

[0047] As an example, the aforementioned implementing entity can first determine the initial multi-level indicator information set of the aforementioned privacy-de-emphasis substation multivariate information set by integrating the experience of multiple experts. Then, using the Pearson correlation coefficient algorithm, correlation analysis is performed on each initial multi-level indicator information set to obtain an indicator correlation value set. Finally, any one of the initial multi-level indicator information pairs whose corresponding indicator correlation value is greater than or equal to a preset correlation threshold is removed from the aforementioned initial multi-level indicator information set to obtain the associated prediction indicator information set.

[0048] Step 104: Based on the multivariate information set of privacy-deprived substations, determine the indicator entropy weight value set and indicator risk weight value set of the associated prediction indicator information set.

[0049] In some embodiments, the aforementioned executing entity can determine the set of index entropy weight values ​​and the set of index risk weight values ​​for the aforementioned associated prediction index information set based on the aforementioned privacy-de-identified substation multivariate information set. The index entropy weight values ​​in the aforementioned set of index entropy weight values ​​can characterize the information entropy of the associated prediction index information, i.e., its degree of dispersion. The index risk weight values ​​in the aforementioned set of index risk weight values ​​can characterize the weight information of the associated prediction index information in terms of substation operational risk.

[0050] In some optional implementations of certain embodiments, determining the set of indicator entropy weight values ​​and indicator risk weight values ​​of the aforementioned associated prediction indicator information set based on the aforementioned privacy-deprived substation multivariate information set may include the following steps:

[0051] The first step involves performing multi-level risk identification on the aforementioned privacy-de-identified substation multi-dimensional information set, resulting in a multi-level risk identification information set for the substation. This multi-level risk identification information set can be information identifying hazards to substation safety at the levels of multi-dimensional operational status information set, multi-dimensional collaborative construction information set, and multi-dimensional target value information set. For example, the multi-level risk identification information set may include: substation operational status risk information, substation collaborative construction risk information, and substation target value risk information. This multi-level risk identification can be performed using a fault tree analysis algorithm.

[0052] The second step is to generate a substation risk response information set for the target substation based on the aforementioned multi-level risk identification information set for substations. The substation risk response information in this set can be information on risk solutions used to address the risks identified in the multi-level risk identification information set for substations.

[0053] As an example, the aforementioned implementing entity can first perform risk quantification processing on the multi-level risk identification information set of the substation to construct a set of substation risk objective functions and substation risk constraint functions. The substation risk objective function can be an objective function that minimizes risk loss, minimizes the probability of risk occurrence, and maximizes power supply reliability. The objective function that minimizes risk loss can be (probability of risk occurrence...) Risk loss The objective function is to minimize the probability of risk occurrence (1 - the risk reduction coefficient of risk response information). The objective function to minimize the probability of risk occurrence can be (the probability of risk occurrence). The objective function is to minimize (1 - the risk reduction coefficient of risk response information). The above objective function for maximizing power supply reliability can be... (1 + Risk Response Information's Improvement Coefficient for Power Supply Reliability). The risk reduction coefficient mentioned above can be a coefficient used to measure the degree to which the probability of risk occurrence is reduced after risk avoidance. The risk reduction coefficient mentioned above can be a value between [0, 1], with a larger value indicating a better effect on risk reduction. The risk response information's improvement coefficient for power supply reliability can be a coefficient used to measure the degree to which the substation's power supply reliability is improved after risk avoidance. The risk response information's improvement coefficient for power supply reliability can be a value between [0, 1], with a larger value indicating a better effect on power supply reliability improvement. The substation risk constraint function set mentioned above can include: equipment capacity constraint function, voltage deviation constraint function, substation input value constraint function, and substation environmental protection constraint function. The equipment capacity constraint function mentioned above can be a function where the load is less than or equal to 0.8 times the transformer's rated capacity. The voltage deviation constraint function can be a function where the absolute value of the difference between the actual voltage and the rated voltage is less than or equal to 0.1 times the rated voltage. The aforementioned substation input value constraint function can be a function where the annual input value attribute value (annual investment amount) is less than or equal to the annual budget value attribute value (annual budget). The aforementioned substation environmental protection constraint function can be a function where noise is less than or equal to the nationally stipulated noise level and the power frequency electric field is less than the nationally stipulated power frequency electric field threshold. Secondly, a hybrid algorithm combining non-dominated sorting genetic algorithm and particle swarm optimization is used to solve the aforementioned substation risk objective function and substation risk constraint function set, obtaining the top 5 substation risk response information sets.

[0054] The third step is to determine the substation risk indicator information set based on the aforementioned multi-level risk identification information set for substations. The substation risk indicator information in this set can be key information used for quantitative statistics to characterize the multi-level risk identification information set for substations. This substation risk indicator information set may include, but is not limited to, at least one of the following: substation equipment failure probability, electricity price volatility, uncertainty of substation investment return, and substation network security level.

[0055] The fourth step is to standardize the risk information set corresponding to the aforementioned substation risk indicator information set to obtain a standardized risk indicator information set. This standardization process can involve converting the risk information set corresponding to the aforementioned substation risk indicator information set into dimensionless standardized data.

[0056] The fifth step is to determine the set of risk indicator variation weights and the set of risk indicator conflict weights for the standardized risk indicator information set. The risk indicator variation weights in the variation weights set can be values ​​used to measure the degree of variation in the standardized risk information, reflecting its relative importance; the larger the value, the higher the relative importance. The risk indicator conflict weights in the conflict weights set can be values ​​used to measure the correlation between standardized risk indicators; the larger the value, the lower the conflict between standardized risk indicators. In practice, the implementing entity can perform the following risk weight determination steps for each standardized risk indicator in the standardized risk indicator information set: First, determine the indicator mean and standard deviation of the standardized risk indicator information. Second, determine the ratio of the indicator mean to the indicator standard deviation as the risk indicator variation weight. Then, using the correlation coefficient calculation formula, determine the risk indicator conflict weight set for each standardized risk indicator in the standardized risk indicator information set.

[0057] Step 6: Based on the aforementioned risk indicator variation weight value set and risk indicator conflict weight value set, perform multi-criteria sorting on the aforementioned substation risk response information set to obtain the target substation risk response information. The target substation risk response information can be the substation risk response information with the largest value in the aforementioned substation risk response information set.

[0058] As an example, the aforementioned implementing entity can first determine the cumulative sum of multiple conflict weight values ​​of risk indicators corresponding to each risk indicator variation weight value in the aforementioned risk indicator variation weight value set, and multiply it by the risk indicator variation weight value, as the comprehensive weight value of the risk indicator, thus obtaining the comprehensive weight value set of risk indicators. Secondly, it determines the ratio of each comprehensive weight value of risk indicators to the sum of the comprehensive weight value set of risk indicators, as the risk weight value of the standardized risk indicator information, thus obtaining the risk weight value set. Next, it standardizes the data corresponding to the standardized risk indicator information set included in the aforementioned substation risk response information set, thus obtaining a standardized response information set. Subsequently, it determines the maximum and minimum values ​​of each standardized response information in the standardized response information set as the positive and negative ideal solutions of the standardized risk indicator information. Then, it determines the product of the square of the difference between each standardized response information in the standardized response information set and its corresponding positive and negative ideal solutions, multiplied by the corresponding risk weight value, thus obtaining the positive ideal distance value set and the negative ideal distance value set. Next, the sets of positive and negative ideal distance values ​​are weighted and summed to generate a ranking index for each substation's risk response information, resulting in a ranking index set. Finally, the substation risk response information corresponding to the ranking index with the largest value in the ranking index set is determined as the target substation risk response information.

[0059] The seventh step is to determine the risk weight value set corresponding to the risk response information of the target substation mentioned above as the indicator risk weight value set.

[0060] Step 8: Based on the aforementioned multi-source information set from privacy-de-scarce substations, determine the set of indicator entropy weight values. In practice, the implementing entity can utilize the entropy weight method to determine the set of indicator entropy weight values ​​based on the aforementioned multi-source information set from privacy-de-scarce substations.

[0061] Furthermore, in the process of adopting technical solutions to address the technical problems mentioned in the background, the following technical issues often arise: Because the privacy-de-enhanced substation multi-source information set involves mixed information with multiple attributes, each possessing complexity and uncertainty, this mixed information makes it difficult to accurately predict the substation's state using the privacy-de-enhanced substation multi-source information set. This makes it impossible to accurately understand the comprehensive benefits of the target substation, resulting in low stability. A conventional solution to address these technical problems is to convert different types of mixed information into privacy-de-enhanced substation multi-source information of the same format. Then, using an approximation-ideal-solution sorting algorithm, weights are assigned to the converted information to obtain a set of index entropy weight values. However, this conventional solution still suffers from the following problems: Due to the complexity, high uncertainty, and strong ambiguity of the mixed information, directly converting different types of mixed information into a unified type makes it difficult to completely retain the uncertainty of different types of mixed information, reducing the quality and accuracy of the mixed information. Furthermore, using only one method to determine weights through an approximation-ideal-solution sorting algorithm has limitations, leading to low weight accuracy, low prediction accuracy for the target substation, and reduced stability of the target substation. Considering the shortcomings of conventional solutions, and taking into account the advantages and current state of the index assignment technology possessed by the applicant's research institute partners in this field, we have decided to adopt the following solution:

[0062] Optionally, the determination of the indicator entropy weight value set and indicator risk weight value set of the aforementioned associated prediction indicator information set based on the aforementioned privacy-de-saturated substation multivariate information set may further include the following steps:

[0063] The first step is to determine the product of the transpose of the indicator scoring matrix and the indicator scoring matrix, which serves as the target indicator scoring matrix. This target indicator scoring matrix is ​​the initial weight matrix for the acquired set of related predictive indicator information. This indicator scoring matrix can be a matrix used by experts in the relevant field to evaluate the set of related predictive indicator information based on a 5-point Likert scale.

[0064] The second step is to determine the weight values ​​of the eigenvector corresponding to the largest eigenvalue of the target indicator scoring matrix, which will be used as the first set of weighted values ​​for the associated prediction indicator information set. In practice, the executing entity can first use the characteristic polynomial method to solve for the largest eigenvalue of the target indicator scoring matrix. Then, the values ​​of the eigenvector corresponding to the largest eigenvalue will be determined as the first set of weighted values ​​for the indicator.

[0065] The third step involves performing fuzzy format conversion on the aforementioned privacy-de-enhanced substation multivariate information set to obtain a substation fuzzy number information set. The substation fuzzy number information in this set can represent the membership relationship of the substation fuzzy numbers within the substation fuzzy number information set. In practice, the implementing entity can first perform format grouping processing on the aforementioned privacy-de-enhanced substation multivariate information set to obtain text variable fuzzy number information sets, numerical variable fuzzy information sets, and normal distribution interval fuzzy information sets, which serve as the substation fuzzy number information set. The text variable fuzzy number information set can be obtained by converting the privacy-de-enhanced substation multivariate information representing the degree of text into an intuitionistic fuzzy number information set through a text intuitionistic fuzzy number mapping form. The text intuitionistic fuzzy number mapping form can be a pre-defined form of intuitionistic fuzzy numbers corresponding to different text degree words. For example, the content of the text intuitionistic fuzzy number mapping form could be content where the text degree word is "good" and the intuitionistic fuzzy number is (1, 0, 0). The fuzzy information of the numerical variables in the aforementioned fuzzy information group can be used for cost-type indicators. The fuzzy format conversion can be the difference between the maximum value and the corresponding value of the privacy-removing substation multivariate information, or the ratio of the difference between the maximum and minimum values, as the format conversion of fuzzy membership. For benefit-type indicators, the fuzzy format conversion can be the difference between the corresponding value and the minimum value of the privacy-removing substation multivariate information, or the ratio of the difference between the maximum and minimum values, as the format conversion of fuzzy membership, with a hesitation degree of 0, a non-membership degree of 1, and a format conversion of the difference between membership degrees. The membership degree of the normal distribution interval fuzzy information in the aforementioned fuzzy information group can be the integral value of the probability density function of the privacy-removing substation multivariate information between the corresponding value and the upper limit of the preset satisfaction interval. The non-membership degree can be the integral value of the probability density function between the preset lower limit of dissatisfaction and the corresponding value. The hesitation degree can be the integral value of the probability density function within the preset hesitation degree interval. The aforementioned preset satisfaction interval, preset dissatisfaction interval, and preset hesitation degree interval can all be value ranges of associated predictive indicator information determined by expert experience.

[0066] The fourth step is to determine the substation multivariate fuzzy matrix of the aforementioned privacy-de-suppressed substation multivariate information set and the aforementioned substation fuzzy number information set. The elements in the aforementioned substation multivariate fuzzy matrix can be a vector composed of membership, non-membership, and hesitation degrees included in the substation fuzzy number information set corresponding to the association prediction index information of the privacy-de-suppressed substation multivariate information set.

[0067] The fifth step involves determining the index fuzzy entropy set of the aforementioned associated prediction index information set based on the substation multivariate fuzzy matrix. The index fuzzy entropy in this set can be used to characterize the fuzziness of the associated prediction index information. For example, the executing entity can determine the vector elements of the substation multivariate fuzzy matrix corresponding to each associated prediction index information in the associated prediction index information set, and use the ratio of the square of the absolute value of the difference between 1 and the difference between the membership and non-membership degrees included in the vector element, the sum of the squares of the hesitation degrees, and 2, as the index fuzzy entropy.

[0068] Step 6: Based on the aforementioned set of fuzzy entropy indicators, determine the second set of weighted values ​​for the aforementioned set of correlated prediction indicator information. The second weighted values ​​in this set represent the weights of the correlated prediction indicator information determined by an objective assignment method. For example, the executing entity can first determine the average of the accumulated fuzzy entropies of each indicator in the aforementioned set of fuzzy entropies to generate an average fuzzy entropy set. Then, it determines the sum of the differences between 1 and each average fuzzy entropy to obtain a comprehensive fuzzy value. Finally, it determines the ratio of the difference between 1 and each average fuzzy entropy in the aforementioned set to the comprehensive fuzzy value, using this ratio as the second weighted value, thus obtaining the second set of weighted values.

[0069] Step 7: Based on the aforementioned first indicator weighted weight value set and the aforementioned second indicator weighted weight value set, determine the indicator entropy weight value set. As an example, the executing entity can first perform the following solution steps for each indicator entropy weight value in the aforementioned indicator entropy weight value set: Using a cooperative game theory algorithm, determine the L2 normal form of the first indicator weighted weight value corresponding to the indicator entropy weight value, the second indicator weighted weight value set, the sum of the first weight coefficient variable and the second weight coefficient variable, and the difference between these values, to obtain the first weight coefficient and the second weight coefficient. Then, determine the sum of the product of the first weight coefficient and each first indicator weighted weight value, and the sum of the product of the second weight coefficient and each second indicator weight value, to generate the indicator entropy weight value, thus obtaining the indicator entropy weight value set.

[0070] Step 8: Based on the aforementioned multi-source information set of the privacy-de-scarce substation, determine the indicator risk weight value set, and dynamically adjust the target substation according to the indicator risk weight value set and the indicator entropy weight value set. For specific implementation methods, refer to the implementation methods of indicator risk weight values ​​in some optional implementation methods of certain embodiments following step 104.

[0071] The above-described technical solution and its related content, as an inventive point of this disclosure, solve the technical problem mentioned in the background art: "Due to the complexity, high uncertainty, and strong ambiguity of mixed information, directly converting different types of mixed information into a unified type of information makes it difficult to completely retain the uncertainty of different types of mixed information, reducing the quality and accuracy of mixed information. Furthermore, determining weights through an approximation-ideal-solution sorting algorithm has limitations, leading to low weight accuracy, low prediction accuracy for the target substation, and reduced stability of the target substation." If these factors are addressed, the accuracy of weights and the prediction accuracy for the target substation can be improved, thereby enhancing the stability of the target substation. To achieve this effect, this disclosure first determines the first indicator weight value set through the obtained indicator scoring matrix, and then determines the first weight through subjective weighting. Secondly, the multivariate information set from the privacy-de-suppressed substation is transformed into mixed information in different formats, including textual variables, numerical variables, and normally distributed variables. The membership, non-membership, and hesitation degrees of intuitionistic fuzzy numbers are used to characterize the content of the original mixed information, which can more comprehensively preserve the uncertainty of the mixed information and reduce information loss. Then, the transformed intuitionistic fuzzy matrix is ​​used to determine the more objective weights for the second indicator. Next, a cooperative game model is used to determine the precise weight sets for the first and second indicator weights, accurately determining the combined weight value of the first and second indicator weights as the indicator entropy weight value. Through multiple combination schemes, the obtained indicator entropy weight value can more accurately reflect the authenticity and importance of the indicator, avoiding the limitations of a single assignment method to some extent. Finally, the obtained indicator entropy weight value set and indicator risk weight value set are used to precisely adjust the target substation, improving its stability.

[0072] Step 105: Generate a target weight set of indicators for the associated prediction indicator information set based on the indicator entropy weight set and the indicator risk weight set.

[0073] In some embodiments, the executing entity can generate a target weight value set for the associated prediction indicator information set based on the aforementioned indicator entropy weight value set and the aforementioned indicator risk weight value set. The target weight values ​​in the target weight value set can be weight values ​​obtained by weighted summation of the aforementioned indicator entropy weight value set and the aforementioned indicator risk weight value set. As an example, the executing entity can use a cooperative game theory algorithm to determine the entropy weight coefficients and risk weight coefficients of the aforementioned indicator entropy weight value set and the aforementioned indicator risk weight value set. Then, the sum of the product of the entropy weight coefficient and each indicator entropy weight value, and the product of the risk weight coefficient and each indicator risk weight value, is determined to generate the indicator target weight values, thus obtaining the indicator target weight value set.

[0074] Step 106: Based on the target weight value set and the associated prediction indicator information set, perform state prediction processing on the target substation to obtain substation state prediction information.

[0075] In some embodiments, the executing entity can perform state prediction processing on the target substation based on the target weight value set and the associated prediction indicator information set to obtain substation state prediction information. The substation state prediction information can be information used to evaluate the comprehensive benefits of the target substation. As an example, the executing entity can utilize the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) algorithm to perform state prediction processing on the target substation based on the target weight value set and the associated prediction indicator information set to obtain substation state prediction information.

[0076] In some optional implementations of certain embodiments, the process of performing state prediction processing on the substation based on the target weight value set of the indicators and the associated prediction indicator information set to obtain substation state prediction information may include the following steps:

[0077] The first step involves determining the first and second target weight sets of the aforementioned associated prediction indicator information set based on the substation multivariate fuzzy matrix corresponding to the aforementioned privacy-de-enhanced substation multivariate information set. The first target weight value in the first target weight set can be a vector composed of the maximum membership degree, minimum non-membership degree, 1, and the hesitation degree of the difference between the maximum membership degree and the minimum non-membership degree in the aforementioned substation multivariate fuzzy matrix corresponding to the associated prediction indicator information. Similarly, the second target weight value in the second target weight set can be a vector composed of the minimum membership degree, maximum non-membership degree, 1, and the hesitation degree of the difference between the minimum membership degree and the maximum non-membership degree in the aforementioned substation multivariate fuzzy matrix corresponding to the associated prediction indicator information.

[0078] The second step involves determining, based on the aforementioned set of associated prediction index information, the first target weight value set and the first and second target direction difference sets of the aforementioned first target weight value set and the aforementioned second target weight value set, respectively. The first target direction difference in the first target direction difference set can be the cosine of the angle between the attribute value vector corresponding to the associated prediction index information and the corresponding first target weight value. Similarly, the second target direction difference in the second target direction difference set can be the cosine of the angle between the attribute value vector corresponding to the associated prediction index information and the corresponding second target weight value. This determination can be made using the formula for calculating the cosine of the vector angle.

[0079] The third step involves determining the first and second cosine distance value sets for the first and second target direction difference sets, respectively, based on the aforementioned target target weight value set. The first cosine distance value in the first cosine distance value set represents the cosine similarity to the first target direction difference set. Similarly, the second cosine distance value in the second cosine distance value set represents the cosine similarity to the second target direction difference set. For example, the executing entity can first determine the sum of the products of each first target direction difference set and its corresponding target weight value to obtain the first cosine distance value set. Then, it can determine the sum of the products of each second target direction difference set and its corresponding target weight value to obtain the second cosine distance value set.

[0080] The fourth step involves determining the state prediction value set of the aforementioned associated prediction index information set based on the first cosine distance value set and the second cosine distance value set. The state prediction values ​​in this set characterize the substation's overall benefits and its proximity to the ideal solution. For example, the implementing entity can first determine the sum of each first cosine distance value in the first cosine distance value set and the corresponding second cosine distance value in the second cosine distance value set, as the target cosine distance value set. Then, it can determine the ratio of each first cosine distance value in the first cosine distance value set to the corresponding target cosine distance value in the target cosine distance value set, as the state prediction value set.

[0081] The fifth step involves performing a weighted summation on the aforementioned state prediction numerical set and the aforementioned privacy-de-identified substation multivariate information set to obtain the substation state prediction information. The weights in the weighted summation can be pre-defined weight values.

[0082] Step 107: In response to the determination that the substation state prediction information does not meet the preset state prediction conditions, perform an operational state attribution analysis on the target substation to obtain substation state attribution information.

[0083] In some embodiments, the executing entity may, in response to determining that the substation state prediction information does not meet preset state prediction conditions, perform an operational state attribution analysis on the target substation to obtain substation state attribution information. The substation state attribution information may be the fundamental factors causing the target substation to fail to meet the preset state prediction conditions. The preset state prediction conditions may be pre-set conditions representing ideal comprehensive benefits that the target substation can achieve. In practice, the executing entity may, in response to determining that the substation state prediction information does not meet the preset state prediction conditions, utilize an attribution analysis algorithm to perform an operational state attribution analysis on the target substation to obtain substation state attribution information.

[0084] Furthermore, in the process of adopting technical solutions to address the technical problems mentioned in the background, the following technical issues often arise: there are certain causal relationships among the associated predictive indicator information sets of the target substation. Modifying any associated predictive indicator information will cause significant changes in the corresponding substation data transmission, resulting in uncontrollable adjustments to the substation's dynamics, leading to lower stability of the target substation and a higher damage rate of power equipment. A conventional solution to these technical problems is to use a single-step random walk algorithm to perform operational status attribution analysis on the target substation by de-privacy substation multivariate information sets, obtaining substation status attribution information. However, this conventional solution still suffers from the following problems: due to the complex calling relationships between the associated predictive indicator information sets and the involvement of multiple associated predictive indicator information, a comprehensive and accurate understanding of the causal relationships between the associated predictive indicator information cannot be achieved, resulting in low efficiency and accuracy in attribution localization, prolonged root cause localization time, lower stability of the target substation and equipment, and increased damage rate of the target substation. Considering the shortcomings of conventional solutions, and taking into account the advantages and current state of attribution analysis technology possessed by the applicant's research institute partners in this field, we have decided to adopt the following solution:

[0085] In some optional implementations of certain embodiments, the above-described operational status attribution analysis of the target substation to obtain substation status attribution information may include the following steps:

[0086] The first step, in response to the determination that the aforementioned substation state prediction information does not meet the preset state prediction conditions, constructs a relationship graph based on the preset substation knowledge graph, connecting the privacy-de-saturated substation multivariate information set and the aforementioned associated prediction indicator information set, generating an indicator association causal relationship graph. This indicator association causal relationship graph can be a directed acyclic graph composed of the associated prediction indicator information set and the relationships between the associated prediction indicator information. Nodes in the graph can be associated prediction indicator information, and edges can be the relationships and causal relationships between the associated prediction indicator information extracted from the privacy-de-saturated substation multivariate information set. The weights of the edges in the graph represent the weighted sum of the degree of association and causality. The preset substation knowledge graph can be an existing knowledge graph related to substations. In practice, the privacy-de-saturated substation multivariate information set and the associated prediction indicator information set are input into the causal graph generation model to obtain the indicator association causal relationship graph. In practice, the aforementioned implementing entity may, in response to determining that the aforementioned substation state prediction information does not meet the preset state prediction conditions, use the NOTEARS (DAGs with NO TEARS) algorithm to construct a relationship graph between the privacy-de-privacy substation multivariate information set and the aforementioned associated prediction indicator information set, and generate an indicator association causal relationship graph.

[0087] The second step involves pruning the aforementioned indicator correlation causal relationship graph based on the target weight value set and the associated node connection importance set, resulting in a pruned indicator correlation causal relationship graph. This pruned graph can retain only the correlation prediction indicator information that causes the substation state prediction information of the target substation to fail to meet the preset state prediction conditions. The associated node connection importance in the associated node connection importance set characterizes the importance of the associated node in the indicator correlation causal relationship graph. For example, the executing entity can first determine the set of node connections between each associated node and its connected associated nodes in the indicator correlation causal relationship graph. Secondly, it can determine the ratio of the number of associated prediction indicator information included in the node connection value set to the number of associated prediction indicator information included in the associated prediction indicator information set, which serves as the associated node connection importance set. Then, it can perform a weighted summation of each target weight value in the target weight value set and the corresponding associated node connection importance in the associated node connection importance set, resulting in the associated node importance value set. The weighting in the aforementioned weighted summation can be a weighted summation of weight values ​​dynamically determined based on different scenario information of the target substation, obtained through statistical verification using historical data corresponding to the scenario information, and the target indicator weight value and the importance set of the associated nodes. Next, at least one associated node with an importance value less than a preset importance threshold is selected from the aforementioned causal relationship graph. This preset importance threshold can be a pre-set value used to determine whether to delete a node. For example, the preset importance threshold could be 0.65. Finally, it is determined whether the causal relationship graph after deleting at least one associated node has an incomplete path problem. If so, at least one associated node causing the incomplete path is added back, resulting in a pruned causal relationship graph.

[0088] The third step involves determining the propagation probability matrix for each of the pruned nodes based on the node weights and edge weights of the pruned nodes in the causal relationship graph. Each element in the propagation probability matrix represents the degree of probability of contagion propagation between the pruned nodes. The edge weights represent the weighted sum of the degree of association and causality between two connected pruned nodes. For example, the executing entity can perform the following propagation probability determination steps for each pruned node: First, determine the set of pruned nodes directly connected upstream to the pruned node, as the target pruned node set. Second, determine the sum of the products of each target pruned node in the target pruned node set and the edge weights of the pruned nodes, as the first probability propagation value. Then, for each target-related pruned node in the target-related pruned node set, the product of the weight of the associated edge of the pruned node and the node weight of the corresponding target-related pruned node is determined as the second probability propagation value, thus obtaining the second probability propagation value set. Finally, the ratio of each second probability propagation value in the above second probability propagation value set to the above first probability propagation value is determined as the propagation probability.

[0089] The fourth step involves determining the set of node similarity values ​​for each adjacent node in the aforementioned set of association prediction index information, based on the category label information set. Each adjacent node is defined as a node connected to the associated node by a single edge. Higher similarity values ​​indicate greater similarity between the associated node and its adjacent nodes. The category label information in the aforementioned set can be the physical attribute information of the association prediction index information. This set can include: physical quantity category labels, measurement object category labels, and equipment level category labels. The physical quantity category labels can reflect the physical nature of the associated pruned node. The measurement object category labels can reflect the information of the power equipment to which the associated pruned node belongs. The equipment level category labels can reflect the hierarchical information of the power system to which the associated pruned node is located. As an example, the aforementioned execution entity can perform the following node similarity determination steps for each associated pruned node included in the pruned index association causal relationship graph: In response to determining that the associated pruned node and its adjacent associated nodes belong to the same type of associated index information, the cosine similarity between the associated pruned node and its adjacent associated nodes is determined as the node similarity value. Here, the aforementioned same-type associated index information can refer to the associated pruned node and its adjacent associated nodes being identical in physical quantity category label information, measurement object category label information, and equipment level category label information; all others are considered dissimilar associated index information. In response to determining that the associated pruned node and its adjacent associated nodes belong to dissimilar associated index information, the product of the cosine similarity between the associated pruned node and its adjacent associated nodes and the dissimilar penalty coefficient matrix is ​​determined and normalized, serving as the node similarity value. Here, the aforementioned dissimilar penalty coefficient matrix can be a pre-defined coefficient matrix composed of the physical quantity category label information of the associated pruned node and its adjacent associated nodes. The horizontal and vertical axes of the above heterogeneous penalty coefficient matrix can be voltage, current, power, temperature, time, and state. The elements in the heterogeneous penalty coefficient matrix can represent the preset association probability of different physical quantity category labels.

[0090] Fifth, based on the aforementioned propagation probability matrix and the aforementioned set of node similarity values, a reverse walk is performed on each of the aforementioned pruned nodes to obtain a set of node walk paths. The node walk paths in this set can be paths that originate from an abnormal node and spread between related nodes. For example, the executing entity can first identify anomalies in the pruned indicator association causal relationship graph to obtain a set of abnormal nodes. Secondly, the pruned indicator association causal relationship graph is reversed to obtain a reverse causal relationship graph. The node weights in the reverse causal relationship graph can be obtained by averaging the node similarity values ​​corresponding to each element in the propagation probability matrix and the node similarity values ​​in the aforementioned set of node similarity values, resulting in a weight value for the node walk transition probability. Then, using the random walk algorithm in the improved Node2Vec algorithm, a random walk is performed on the reverse causal relationship graph based on the reverse walk constraint information set and the set of abnormal nodes to obtain a set of node walk paths. The reverse walk constraint information in the aforementioned reverse walk constraint information set can be pre-defined walk constraint information. For example, the aforementioned reverse walk constraint information could be that the starting point of the walk path needs to be from the abnormal node, the downstream node of the abnormal node has been visited, reducing the probability of visit, and the maximum walk depth is 10. In the improved Node2Vec algorithm, the return probability coefficient is set to 0.3, the exploration probability coefficient is set to 2, the walk length can be in the range of [8, 10], and the number of samplings, i.e., the number of times to perform multiple walks for the abnormal node, can be 200.

[0091] Step 6: Based on the aforementioned node walk path set, determine the root cause value set of each associated pruned node. The root cause values ​​in this set represent the probability that the associated node is causing the target substation to fail to meet the preset state prediction conditions. As an example, the executing entity can perform the following multiplication process for each associated pruned node in the causal relationship diagram of the pruned indicators: First, determine the product of the propagation probability set of each associated pruned node in at least one node walk path corresponding to the associated pruned node and the corresponding node weight, and the cumulative product of the time decay factor, as the walk path weight value, to obtain the walk path weight value set. The time decay factor represents the time delay caused by the propagation duration. The time decay factor can be an exponential function with base e and the exponent being the negative of the product of the decay coefficient and the propagation duration of the node walk path. Secondly, the ratio of each path weight in the path weight set to the average degree of the aforementioned pruned nodes is determined as the path contribution value of the path, thus obtaining the path contribution value set of the path. The average degree can be the mean of the out-degree and in-degree of the aforementioned pruned nodes. Next, the logarithm of the sum of the number of associated edges of the path of at least one node (base 10) and 1 is determined as the path depth contribution value of the path. Then, the node penalty contribution is obtained by multiplying the product of the out-degree of the aforementioned pruned node (base e), the ratio of the node's in-degree to the sum of 1 and 1, and a suppression factor. The suppression factor characterizes the inhibitory effect on key pruned nodes with many connected edges, since a pruned node with many connected edges is not necessarily a root cause node. Finally, the combined multiplication of the path contribution value set, the path depth contribution value, and the node penalty contribution value yields the node root cause value.

[0092] Step 7: Dynamically adjust the target substation based on the root cause value set. In practice, at least one associated pruned node with a root cause value greater than or equal to a preset root cause threshold is selected from the root cause value set and used as anomaly root cause indicator information. Then, the target substation is dynamically adjusted according to the aforementioned anomaly root cause indicator information set. This dynamic adjustment could be based on the substation's status attribution information corresponding to aging power equipment, or by purchasing new power equipment.

[0093] The above-mentioned technical solution and related content, as an inventive point of this disclosure, solve the technical problem mentioned in the background: "Due to the complex calling relationships between the information sets of related prediction indicators and the involvement of multiple related prediction indicator information, it is impossible to fully and accurately understand the causal relationship between the related prediction indicator information, resulting in low efficiency and accuracy of attribution localization, prolonged root cause localization time, low stability of the target substation and equipment, and increased damage rate of the target substation." If the above factors are solved, the accuracy of weights and the prediction accuracy of the target substation can be improved, thereby increasing the stability of the target substation. To achieve this effect, this disclosure first constructs a causal relationship graph using the NOTEARS algorithm and a preset substation knowledge graph. By introducing substation domain constraints, the accuracy of constructing the indicator association causal relationship graph can be improved, and it can better reflect real substation scenarios. Second, pruning the indicator association causal relationship graph can remove duplicate, redundant, and useless nodes, balancing pruning efficiency and causal chain integrity, reducing the amount of computational data for subsequent attribution localization, and improving the efficiency of attribution analysis. Then, the propagation probability and node similarity values ​​of the associated pruned nodes are determined. Based on the propagation probability and node similarity values, the node traversal path set is determined. This considers the upstream and downstream causal relationships in the causal relationship graph of the pruned indicators and the similarity constraints of physical labels, facilitating the accurate determination of the propagation capabilities of abnormal nodes. This makes the analysis more consistent with the real scenario of the target substation and allows for subsequent accurate determination of traversal paths for attribution analysis. Next, based on the node traversal path set, the root cause values ​​of each associated pruned node are determined. By considering three dimensions—path contribution, depth contribution, and node penalty contribution for inaccurate causal attribution analysis caused by suppressing highly connected nodes—the accuracy of determining the root cause values ​​of each associated pruned node can be improved, avoiding the bias caused by a single aspect and better reflecting the target substation scenario. Finally, dynamic adjustments to the target substation using the node root cause values ​​can shorten the attribution location time, improve the stability and equipment stability of the target substation, enhance its safety, and reduce its damage rate.

[0094] Step 108: Dynamically adjust the target substation based on the substation status attribution information.

[0095] In some embodiments, the implementing entity can dynamically adjust the target substation based on the target substation status attribution information. This dynamic adjustment can be based on problems identified in the substation status attribution information regarding any aspect of the target substation, including operational status, target value, or collaborative construction, and adjustments can be made targeting the problematic aspects. For example, if the problem lies in operational status, the granularity of fault detection for substation equipment could be increased, or new power equipment could be replaced.

[0096] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a substation dynamic adjustment device, which are similar to... Figure 1 Corresponding to the method embodiments shown, this substation dynamic adjustment device can be specifically applied to various electronic devices.

[0097] like Figure 2 As shown, a substation dynamic adjustment device 200 includes: an acquisition unit 201, a privacy removal unit 202, a first determination unit 203, a second determination unit 204, a generation unit 205, a state prediction unit 206, an attribution analysis unit 207, and a dynamic adjustment unit 208. The acquisition unit 201 is configured to acquire a multi-dimensional operating status information set, a multi-dimensional collaborative construction information set, and a multi-dimensional target value information set of the target substation. The privacy removal unit 202 is configured to perform privacy removal processing on the aforementioned multi-dimensional operating status information set, the aforementioned multi-dimensional collaborative construction information set, and the aforementioned multi-dimensional target value information set to obtain a privacy-removed substation multi-dimensional information set. The first determination unit 203 is configured to determine a set of associated prediction index information for the target substation based on the aforementioned privacy-removed substation multi-dimensional information set. The second determination unit 204 is configured to determine the index entropy weight value set and the index risk weight value set of the aforementioned associated prediction index information set based on the aforementioned privacy-removed substation multi-dimensional information set. The generation unit 205 is configured to generate a target weight value set for the associated prediction indicator information set based on the aforementioned indicator entropy weight value set and the aforementioned indicator risk weight value set. The state prediction unit 206 is configured to perform state prediction processing on the aforementioned target substation based on the aforementioned indicator target weight value set and the aforementioned associated prediction indicator information set to obtain substation state prediction information. The attribution analysis unit 207 is configured to perform operational state attribution analysis on the aforementioned target substation in response to determining that the aforementioned substation state prediction information does not meet preset state prediction conditions, to obtain substation state attribution information. The dynamic adjustment unit 208 is configured to dynamically adjust the aforementioned target substation based on the aforementioned substation state attribution information.

[0098] It is understandable that the units described in the substation dynamic adjustment device 200 are related to the reference. Figure 1 The steps described in the method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the substation dynamic adjustment device 200 and the units contained therein, and will not be repeated here.

[0099] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device (e.g., an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0100] like Figure 3 As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0101] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0102] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0103] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0104] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0105] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the contents of steps 101-108.

[0106] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0107] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0108] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including an acquisition unit, a privacy-removing unit, a first determination unit, a second determination unit, a generation unit, a state prediction unit, an attribution analysis unit, and a dynamic adjustment unit. The names of these units do not necessarily limit the specific unit; for example, the acquisition unit may also be described as "a unit that acquires a multi-dimensional operating status information set, a multi-dimensional collaborative construction information set, and a multi-dimensional target value information set of a target substation."

[0109] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0110] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A dynamic adjustment method for a substation, comprising: Acquire a multi-dimensional set of operational status information, a multi-dimensional set of collaborative construction information, and a multi-dimensional set of target value information for the target substation; The privacy-de-identifying substation multi-dimensional information set is obtained by performing privacy-de-identifying multi-dimensional information set on the multi-dimensional operational status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set. Based on the privacy-de-scarce substation multivariate information set, determine the associated prediction indicator information set for the target substation; Based on the privacy-deprived substation multivariate information set, determine the indicator entropy weight value set and indicator risk weight value set of the associated prediction indicator information set; Based on the set of indicator entropy weight values ​​and the set of indicator risk weight values, generate a set of target weight values ​​for the associated prediction indicator information set; Based on the target weight value set of the indicators and the associated prediction indicator information set, state prediction processing is performed on the target substation to obtain substation state prediction information, including: determining a first target weight value set and a second target weight value set of the associated prediction indicator information set based on the substation multivariate fuzzy matrix corresponding to the privacy-removed substation multivariate information set; determining a first target direction difference set and a second target direction difference set of the first target weight value set and the second target weight value set based on the associated prediction indicator information set; determining a first cosine distance value set and a second cosine distance value set of the first target direction difference set and the second target direction difference set based on the target weight value set of the indicators; determining a state prediction value set of the associated prediction indicator information set based on the first cosine distance value set and the second cosine distance value set; and performing a weighted summation process on the state prediction value set and the privacy-removed substation multivariate information set to obtain substation state prediction information. In response to determining that the substation state prediction information does not meet the preset state prediction conditions, an operational state attribution analysis is performed on the target substation to obtain substation state attribution information. The target substation is dynamically adjusted based on the substation status attribution information.

2. The method according to claim 1, wherein, The privacy-de-identifying process is performed on the multi-dimensional operational status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set to obtain a privacy-de-identified substation multi-dimensional information set, including: Unstructured information is filtered out from the multi-dimensional operational status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set to obtain a multi-dimensional substation unstructured information set; Based on a pre-defined substation knowledge graph, named entity recognition is performed on the multi-dimensional substation unstructured information set to obtain a substation entity information set. Contextual semantic features are extracted from the substation entity information set to obtain a substation semantic feature vector set; Determine the entity semantic annotation information set for each substation entity information in the substation entity information set; The entity semantic annotation information set and the substation semantic feature vector set are subjected to feature embedding processing to obtain the substation text embedding feature vector set; The text embedding feature vector set of the substation is input into the privacy recognition model to obtain unstructured privacy data; The unstructured privacy data is de-privacy processed to obtain a privacy-deprivacy substation multi-source information set.

3. The method according to claim 2, wherein, The privacy-de-identifying process is performed on the multi-dimensional operational status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set to obtain a privacy-de-identified substation multi-dimensional information set, including: The substation unstructured information set is obtained by removing the multi-dimensional substation unstructured information set from the multi-dimensional operational status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set; Determine the structured information entropy set and the structured discrete entropy set of the substation's structured information set; Based on the structured information entropy set and the structured discrete entropy set, a structured privacy degree set is generated for the structured information set of the substation. Based on the structured privacy degree set, the structured information set of the substation is subjected to privacy theme clustering processing to obtain a structured privacy data cluster set and a structured non-privacy data cluster set; Determine the cluster association information set of the structured privacy data cluster set and the structured non-privacy data cluster set; Based on the cluster association information set, the structured privacy data cluster set and the structured non-privacy data cluster set are subjected to cluster update processing to obtain the updated structured privacy data cluster set; The updated structured privacy data clusters are de-privacy processed to obtain a privacy-deprivacy substation multi-dimensional information set.

4. The method according to claim 1, wherein, The step of determining the set of indicator entropy weight values ​​and indicator risk weight values ​​of the associated prediction indicator information set based on the privacy-de-saturated substation multivariate information set includes: Multi-level risk identification is performed on the privacy-deprivation substation multi-dimensional information set to obtain a multi-level risk identification information set for the substation. Based on the multi-level risk identification information set of the substation, a substation risk response information set for the target substation is generated; Based on the multi-level risk identification information set of the substation, determine the risk indicator information set of the substation; The risk information set corresponding to the substation risk indicator information set is standardized to obtain a standardized risk indicator information set. Determine the set of risk indicator variation weight values ​​and the set of risk indicator conflict weight values ​​for the standardized risk indicator information set; Based on the risk indicator variation weight value set and the risk indicator conflict weight value set, the substation risk response information set is sorted by multiple criteria to obtain the target substation risk response information. The risk weight value set corresponding to the risk response information of the target substation is determined as the indicator risk weight value set; Based on the multivariate information set of the privacy-de-scarce substation, determine the set of indicator entropy weight values.

5. A dynamic adjustment device for a substation, comprising: The acquisition unit is configured to acquire a set of multi-dimensional operating status information, a set of multi-dimensional collaborative construction information, and a set of multi-dimensional target value information of the target substation; The privacy-removing unit is configured to perform privacy-removing processing on the multi-dimensional operating status information set, the multi-dimensional collaborative construction information set, and the multi-dimensional target value information set to obtain a privacy-removing substation multi-dimensional information set. The first determining unit is configured to determine a set of associated predictive indicator information for the target substation based on the privacy-de-scarce substation multivariate information set. The second determining unit is configured to determine the set of indicator entropy weight values ​​and the set of indicator risk weight values ​​of the associated prediction indicator information set based on the privacy-deprived substation multivariate information set. The generation unit is configured to generate a target weight value set for the associated prediction indicator information set based on the indicator entropy weight value set and the indicator risk weight value set. A state prediction unit is configured to perform state prediction processing on the target substation based on the target weight value set of the indicators and the associated prediction indicator information set to obtain substation state prediction information. This includes: determining a first target weight value set and a second target weight value set of the associated prediction indicator information set based on the substation multivariate fuzzy matrix corresponding to the privacy-removed substation multivariate information set; determining a first target direction difference set and a second target direction difference set of the first target weight value set and the second target weight value set, respectively, based on the associated prediction indicator information set; determining a first cosine distance value set and a second cosine distance value set of the first target direction difference set and the second target direction difference set, respectively, based on the target weight value set of the indicators; determining a state prediction value set of the associated prediction indicator information set based on the first cosine distance value set and the second cosine distance value set; and performing a weighted summation process on the state prediction value set and the privacy-removed substation multivariate information set to obtain the substation state prediction information. The attribution analysis unit is configured to perform an operational state attribution analysis on the target substation in response to determining that the substation state prediction information does not meet the preset state prediction conditions, and to obtain substation state attribution information. The dynamic adjustment unit is configured to dynamically adjust the target substation based on the substation status attribution information.

6. An electronic device, comprising: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-4.

7. A computer-readable medium having a computer program stored thereon, wherein, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-4.

Images