Resident power supply abnormity operation and maintenance method and related device

Abstract

The invention provides a resident power supply abnormity operation and maintenance method and a related device, and belongs to the technical field of power system operation and maintenance. According to the method, a basic item set, a combined item set and a situation item set are constructed based on fault positioning information and an emotion vocabulary set; based on the basic item set, the combined item set and the situation item set, determining a frequent item set and a strong association rule; mapping the frequent item set and the strong association rule into an emotion level label to obtain emotion level label data; based on the real-time resident area environment data and user historical feedback data, scene correction parameters are obtained; after the identity of the user is judged by adopting the user identity supplementary data, adjusting the emotion level label data based on the scene correction parameters to obtain adjusted emotion level label data; and comparing the adjusted emotion level label data with the historical event data to obtain a comparison result, and setting an operation and maintenance strategy based on the comparison result. The problem that the operation and maintenance efficiency is not high when power supply of residents is abnormal is solved.

Description

Technical Field

[0001] This invention belongs to the field of power system operation and maintenance technology, specifically relating to a method and related device for abnormal operation and maintenance of residential power supply. Background Technology

[0002] With the expansion and intelligent upgrading of the power system, power outages in residential areas are causing increasingly diverse and complex power supply anomalies. Voltage and frequency anomalies leading to power outages and subsequent power interruptions are becoming more serious, posing challenges to operation and maintenance work. At the same time, users have higher expectations for power supply services. When faced with power outages, users not only want quick resolution but also timely and effective feedback and emotional support. Against this backdrop, the power grid operation and maintenance system needs to integrate massive amounts of multi-source power outage fault data from the power system, efficiently fuse multi-modal data, achieve rapid verification and response decisions for power outage events, and build a reliable system for rapid emotional feedback and communication response for users. However, the current power grid operation and maintenance system still faces the following problems: 1. Regarding user feedback, fault identification, and emotion level determination: The user-facing response modules in existing power grid operation and maintenance systems cannot determine user emotions and intentions, let alone provide personalized feedback services based on user emotional states. They also lack knowledge reasoning capabilities. Faced with diverse natural language text from users, they cannot accurately map it into clear and actionable demand identification results. As a result, the intelligent scheduling of fault handling processes lacks a solid semantic understanding foundation, making it difficult to build an intelligent response closed loop effectively driven by user needs, which greatly limits the improvement of operation and maintenance service quality.

[0003] 2. User-oriented feedback response generation: Existing power grid operation and maintenance systems are extremely lacking in the means of verifying power outage accidents and the methods of generating feedback information. It is difficult to build a cross-modal mapping mechanism for historical fault feedback information, resulting in a technical gap between the semantic comparison of unstructured user descriptions and historical records in the database. Furthermore, it is impossible to conduct regular analysis of historical power outage events in the corresponding area and to effectively identify the types of power outage accidents and conduct in-depth analysis of their underlying causes.

[0004] 3. Regarding the relevance and timeliness of feedback responses in power grid operation and maintenance systems: Firstly, existing power grid operation and maintenance systems struggle to semantically deconstruct and differentiate patterns in feedback information from planned and unplanned power outages, leading to a misalignment between event types and response strategies. Furthermore, a quantitative assessment method for the timeliness of user feedback has not yet been established, making it impossible to intelligently sort high-priority and high-urgency user feedback information using time-series characteristics (such as feedback time intervals and fault duration) weights. In addition, existing power grid operation and maintenance systems lack a response strategy generation mechanism driven by electricity environment awareness, making it difficult to combine environmental variables such as user geographical attributes, load characteristics, and electricity consumption periods to generate scenario-adaptive differentiated response content. Summary of the Invention

[0005] The purpose of this invention is to provide a method and related apparatus for the operation and maintenance of residential power supply anomalies, in order to solve the problem of low operation and maintenance efficiency in the face of residential power supply anomalies in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for maintaining and repairing abnormal power supply in residential areas, comprising the following steps: Obtain information on power outages reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data; Emotion recognition is performed on power outage information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words; Based on fault location information and sentiment vocabulary, basic itemsets, combined itemsets, and contextual itemsets are constructed. Based on basic itemsets, combined itemsets, and contextual itemsets, frequent itemsets and strong association rules are determined. Frequent itemsets and strong association rules are mapped to sentiment level labels to obtain sentiment level label data. Based on real-time residential area environmental data and historical user feedback data, scene correction parameters are obtained; After using supplementary user identity data to determine the user's identity, the emotion level label data is adjusted based on the scenario correction parameters to obtain the adjusted emotion level label data. The adjusted emotion level label data and historical event data are compared to obtain the comparison results, and operation and maintenance strategies are set based on the comparison results.

[0007] A further improvement of the present invention is that the emotion recognition of the power outage fault information reported by residents and the historical user feedback data to obtain fault location information and a set of emotional words is specifically: using the BERT (Bidirectional Encoder Representations from Transformers) model to perform emotion recognition on the power outage fault information reported by residents and the historical user feedback data to obtain fault location information and a set of emotional words.

[0008] A further improvement of this invention is that, before constructing the basic itemset, combined itemset, and contextual itemset based on the fault location information and the sentiment vocabulary set, stop word filtering and part-of-speech normalization are performed on the fault location information and the sentiment vocabulary set.

[0009] A further improvement of this invention is that the determination of frequent itemsets and strong association rules based on basic itemsets, combined itemsets, and contextual itemsets specifically involves: From the basic itemset, combined itemset, and contextual itemset, filter out itemsets that are not frequent enough and retain only the frequently occurring itemsets to obtain frequent itemsets; Based on frequent itemsets, the confidence level is determined, and then the strong association rules are determined based on the confidence level.

[0010] A further improvement of this invention is that the mapping of frequent itemsets and strong association rules to sentiment level labels to obtain sentiment level label data specifically includes: Map frequent itemsets and strong association rules to predicted sentiment levels; Based on the predicted values ​​of the emotion levels, emotion level label data is obtained.

[0011] A further improvement of this invention is that the formula for calculating the predicted value of the emotion level is:

[0012] in, This is a predicted value for mood level. For the first i The weights of each feature The result is the normalization of the features.

[0013] A further improvement of this invention lies in the comparison of the adjusted emotion level label data and historical event data to obtain a comparison result, and the setting of an operation and maintenance strategy based on the comparison result, specifically as follows: If the user's emotional level is low, provide a standardized and concise response. If the user's emotional level is high, respond with empathetic expressions and report to the dispatch department for expedited processing.

[0014] Secondly, the present invention provides a residential power supply anomaly maintenance system, comprising: The data acquisition module is used to acquire power outage fault information reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data; The emotion recognition module is used to perform emotion recognition on power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words. The itemset construction module is used to construct basic itemsets, composite itemsets, and contextual itemsets based on fault location information and sentiment vocabulary sets. The rule determination module is used to determine frequent itemsets and strong association rules based on basic itemsets, combined itemsets, and contextual itemsets. The mapping module is used to map frequent itemsets and strong association rules to sentiment level labels, thereby obtaining sentiment level label data. The scene correction parameter determination module is used to obtain scene correction parameters based on real-time residential area environmental data and user historical feedback data; The adjustment module is used to determine the user's identity using supplementary user identity data, and then adjust the emotion level label data based on the scenario correction parameters to obtain the adjusted emotion level label data. The operations and maintenance module is used to compare the adjusted emotion level label data with historical event data, obtain the comparison results, and set operations and maintenance strategies based on the comparison results.

[0015] Thirdly, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method for abnormal operation and maintenance of residential power supply.

[0016] Fourthly, the present invention provides a storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of the residential power supply anomaly maintenance method described above.

[0017] Compared with the prior art, the present invention has the following beneficial effects: The proposed method for maintaining abnormal power supply in residential areas involves two aspects. First, it performs emotion recognition on power outage reports from residents and historical user feedback data to obtain fault location information and a set of emotional terms. This process transforms unstructured resident requests into quantifiable maintenance priority indicators, thereby shortening the maintenance cycle. Second, after identifying users using supplementary user identity data, it adjusts the emotion level label data based on scenario correction parameters, resulting in adjusted emotion level label data. This process makes the data more closely aligned with actual maintenance needs. Furthermore, it compares the adjusted emotion level label data with historical event data to obtain comparison results. Based on these results, maintenance strategies are set. This approach not only reduces the subjectivity and redundancy of maintenance but also minimizes judgment biases among maintenance personnel. Attached Figure Description

[0018] Figure 1 This is a flowchart of the residential power supply anomaly operation and maintenance method of the present invention; Figure 2 This is a schematic diagram of the residential power supply anomaly maintenance system of the present invention; Figure 3 This is a flowchart of the residential power supply anomaly operation and maintenance method in Embodiment 4 of the present invention; Figure 4 This is a flowchart of emotion recognition in Embodiment 4 of the present invention; Figure 5 This is a flowchart of the emotion level calculation in Embodiment 4 of the present invention; Figure 6This is a schematic diagram of the intelligent response system in Embodiment 4 of the present invention; Figure 7 This is a schematic diagram of the structure of the electronic device of the present invention. Detailed Implementation

[0019] To further understand the content of this invention, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments are merely illustrative and not limiting of the invention.

[0020] Example 1: The flowchart of the residential power supply anomaly maintenance method of this invention is as follows: Figure 1 As shown, the residential power supply anomaly maintenance method of the present invention includes the following steps: S1. Obtain power outage information reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data; S2. Perform emotion recognition on power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words; S3. Based on fault location information and sentiment vocabulary set, construct basic itemsets, combined itemsets, and contextual itemsets; S4. Based on basic itemsets, combined itemsets, and contextual itemsets, determine frequent itemsets and strong association rules; S5. Map frequent itemsets and strong association rules to sentiment level labels to obtain sentiment level label data; S6. Based on real-time residential area environmental data and historical user feedback data, obtain scene correction parameters; S7. After using supplementary user identity data to determine the user's identity, adjust the emotion level label data based on the scenario correction parameters to obtain the adjusted emotion level label data; S8. Compare the adjusted emotion level label data with historical event data to obtain the comparison results, and set operation and maintenance strategies based on the comparison results.

[0021] Example 2: A schematic diagram of the residential power supply anomaly maintenance system of this invention is shown below. Figure 2 As shown, the residential power supply anomaly maintenance system of the present invention includes: The data acquisition module is used to acquire power outage fault information reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data; The emotion recognition module is used to perform emotion recognition on power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words. The itemset construction module is used to construct basic itemsets, composite itemsets, and contextual itemsets based on fault location information and sentiment vocabulary sets. The rule determination module is used to determine frequent itemsets and strong association rules based on basic itemsets, combined itemsets, and contextual itemsets. The mapping module is used to map frequent itemsets and strong association rules to sentiment level labels, thereby obtaining sentiment level label data. The scene correction parameter determination module is used to obtain scene correction parameters based on real-time residential area environmental data and user historical feedback data; The adjustment module is used to determine the user's identity using supplementary user identity data, and then adjust the emotion level label data based on the scenario correction parameters to obtain the adjusted emotion level label data. The operations and maintenance module is used to compare the adjusted emotion level label data with historical event data, obtain the comparison results, and set operations and maintenance strategies based on the comparison results.

[0022] Example 3: The method for maintaining abnormal power supply in residential buildings according to the present invention includes the following steps: S1. Obtain power outage information reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data.

[0023] S2. Perform emotion recognition on power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words.

[0024] In this step, emotion recognition is performed on the power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words. Specifically, the BERT model is used to perform emotion recognition on the power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words.

[0025] S3. Based on fault location information and sentiment vocabulary set, construct basic itemsets, combined itemsets and contextual itemsets.

[0026] Before constructing the basic itemset, combined itemset, and contextual itemset based on the fault location information and the sentiment vocabulary set in this step, stop word filtering and part-of-speech normalization are performed on the fault location information and the sentiment vocabulary set.

[0027] S4. Based on basic itemsets, combined itemsets, and contextual itemsets, determine frequent itemsets and strong association rules.

[0028] This step determines frequent itemsets and strong association rules based on basic itemsets, combined itemsets, and contextual itemsets, specifically as follows: From the basic itemset, combined itemset, and contextual itemset, filter out itemsets that are not frequent enough and retain only the frequently occurring itemsets to obtain frequent itemsets; Based on frequent itemsets, the confidence level is determined, and then the strong association rules are determined based on the confidence level.

[0029] S5. Map frequent itemsets and strong association rules to sentiment level labels to obtain sentiment level label data.

[0030] This step maps frequent itemsets and strong association rules to sentiment level labels, obtaining sentiment level label data, specifically including: Map frequent itemsets and strong association rules to predicted sentiment levels; Based on the predicted values ​​of the emotion levels, emotion level label data is obtained.

[0031] The formula for calculating the predicted value of mood level is:

[0032] in, This is a predicted value for mood level. For the first i The weights of each feature The result is the normalization of the features.

[0033] S6. Based on the residential area environmental data and resident identity data, adjust the emotion level label data to obtain the adjusted emotion level label data.

[0034] S7. Based on the adjusted sentiment level label data, set up operation and maintenance strategies.

[0035] This step compares the adjusted emotion level label data with historical event data to obtain comparison results. Based on these results, an operation and maintenance strategy is set, specifically as follows: If the user's emotional level is low, provide a standardized and concise response. If the user's emotional level is high, respond with empathetic expressions and report to the dispatch department for expedited processing.

[0036] Example 4: The flowchart of the residential power supply anomaly maintenance method of this invention is as follows: Figure 3 As shown, the method of the present invention is specifically implemented through an intelligent agent. The method of the present invention for abnormal operation and maintenance of residential power supply includes the following steps: Step 1: Local power supply workers receive power outage information from residents (also called users). At the outset of the entire operation and maintenance methodology, when users report power outages, the intelligent agent's permission configuration will maintain equal collaboration with frontline power supply workers. The intelligent agent will, with the identity and permissions of a frontline power supply worker, comprehensively acquire and monitor in real-time messages from various social groups (such as WeChat and QQ) within its jurisdiction. This mechanism aims to ensure that the intelligent agent can comprehensively and in real-time capture fault feedback information from the user side.

[0037] Grassroots power supply workers also have the authority to view user feedback information in real time, thus forming an information exchange and permission collaboration mechanism between the intelligent agent and the human maintenance personnel. This permission design ensures that the intelligent agent can obtain comprehensive and real-time fault feedback information from the user side.

[0038] Ensuring that grassroots power supply workers have a synchronized grasp of user feedback, this information sharing and authorization coordination mechanism provides a solid information foundation and authorization guarantee for subsequent fault response and handling. It constructs an integrated information flow between intelligent agents and human operation and maintenance entities, thereby providing a solid information foundation and authorization guarantee for subsequent fault operation and maintenance and handling.

[0039] Step 2: Emotion Recognition Emotion recognition was performed on unprocessed text feedback from users (power outage information reported by residents), resulting in two key types of structured data. The flowchart for emotion recognition is as follows. Figure 4 As shown.

[0040] The first type of data is fault location information, which carries key technical parameters such as fault type, time of occurrence, and spatial location. The second type of data is a set of emotional words used to characterize the emotional tendencies and polarities contained in user feedback.

[0041] Among these, fault location information serves as the core basis, being input into the subsequent intelligent response system for comparative analysis of historical fault location information and accurate verification of user identity. The set of emotional vocabulary is quantified using subsequent emotion level calculation methods, aiming to achieve an accurate assessment of the urgency of users' electricity needs.

[0042] Two types of multi-dimensional data (two key types of structured data) support the intelligent response system in constructing a closed-loop response process and optimizing decisions.

[0043] Step 3: Emotion Level Calculation Figure 5 The flowchart for calculating the emotion level is shown below, along with the specific implementation steps: A. Itemset Construction Feature engineering (preprocessing) is performed on the fault location information and sentiment lexicon. Feature engineering includes stop word filtering and part-of-speech normalization. The stop word filtering and part-of-speech normalization are described in detail below: Stop word filtering uses a pre-defined stop word list and regular expression matching mechanism to accurately remove redundant affixes, meaningless conjunctions, punctuation marks, and other non-critical information from target words, thereby effectively reducing data noise.

[0044] Part-of-speech normalization, using lemmatization and synonym induction techniques, maps words with the same semantic core to standard terminology in the field of power faults. For example, it normalizes synonyms such as "tripping" and "damage to the circuit breaker," thereby significantly enhancing the consistency and computability of feature representation.

[0045] During the itemset construction process, the technology sequentially constructs the basic itemset, composite itemset, and contextual itemset, aiming to establish a structured data foundation covering multi-dimensional semantics. The construction of the basic itemset, composite itemset, and contextual itemset is completed in sequence. The following provides a detailed explanation of the basic itemset, composite itemset, and contextual itemset: The basic itemset takes the target word after feature processing as the object and represents an independent semantic unit, such as "power outage", "anxiety" and "anxiety".

[0046] Composite itemsets structurally associate multiple target words that co-occur in the base itemset to form itemset entities with compound semantics, such as "power outage - anxiety" and "tripping - power outage - anxiety".

[0047] Contextual itemsets focus on integrating contextual elements such as the location and time of the fault after processing, such as "late at night", "early morning" and "peak electricity consumption period".

[0048] Given that users’ tolerance and emotional response to power outages vary at different times, this type of itemset is given the highest weight in subsequent support and association rule calculations. This hierarchical construction, from features to association patterns to spatiotemporal context, provides strong structured data support for subsequent association rule mining.

[0049] B. Support Calculation and Association Rule Generation Support calculation aims to filter statistically significant association patterns from the various itemsets (basic itemsets, combined itemsets, and contextual itemsets) generated in step A. Since itemsets or rules with low support may appear randomly and lack universality in the dataset, a minimum support threshold of 0.2 is set to filter itemsets with insufficient frequency, retaining only those that frequently appear in the dataset; these itemsets are called frequent itemsets. For example, the support of the itemset {"Peak electricity usage", "Appliance malfunction"} = number of responses containing this combination / total number of responses. Once the support of an itemset exceeds the set threshold, it is designated as a frequent itemset. Frequent itemsets receive higher weights to increase their contribution to subsequent sentiment level mapping. The contribution is calculated using the following formula:

[0050] in, Indicates the degree of contribution.

[0051] Association rules are in the form of The implication expression, where and These are disjoint itemsets. The strength of association rules is usually measured by support and confidence. Support indicates how frequently a rule appears in the dataset, and confidence is used to determine... In containing The frequency of occurrence in transactions is calculated using the following formula:

[0052] in, Indicates the confidence level.

[0053] Generate based on frequent itemsets. First, for each frequent itemset... Divide it into two non-empty subsets. and , to obtain the form of The rules are then calculated. The method sets a minimum confidence threshold of 0.7, retaining rules with a confidence level greater than or equal to this threshold, defining them as strongly associated rules. Each frequent itemset... At most can produce There are several association rules. When a set of items matching a strong association rule appears in user feedback, the system will prioritize resources for emotional support and maintenance.

[0054] C. Emotional ranking mapping and classification decision-making This step transforms the quantitative analysis results (frequent itemsets and strong association rules) into discrete emotion level labels to drive intelligent response decisions. The mapping rule base is constructed by classifying and storing the selected frequent itemsets and strong association rule itemsets according to emotion levels (e.g., 1-5 levels), thereby building a mapping rule base of "feature combination → emotion level".

[0055] Next, a linear weighted model mapping is performed. Based on the itemset construction results and association rule construction results, differentiated weights are assigned to different types of features. Predictive value of mood level A linear weighted model is used to map features to levels, predicting the sentiment level. The calculation formula is:

[0056] in, This is a predicted value for mood level. For the first i The weights of each feature (fault location information and sentiment vocabulary set), The normalized results are obtained from the feature fault location information and the sentiment vocabulary set. Finally, a preset threshold division mechanism (e.g.) is used. The value range corresponds to a five-level scale, and the continuous mapping results will be... The data is converted into discrete emotion level labels (emotion level label data), and the result is output to the subsequent intelligent response system.

[0057] Step 4: Adjust the emotion level label data and set up operation and maintenance strategies. Step 4 is specifically implemented through an intelligent response system. Figure 6 The following is a schematic diagram of an intelligent response system (hereinafter referred to as the system). Figure 6 Detailed explanation: The data integration module in the intelligent response system processes data in four parts: emotion level label data obtained in step 3, user identity supplementary data, electricity consumption area environmental data (also called real-time residential area environmental data), and historical user feedback data (also called user historical feedback data).

[0058] During the intelligent agent's response process, the data integration module receives basic itemsets, combined itemsets, contextual itemsets, and emotion level labels. It then combines these with electricity environment data (residential area environment data) and fault feedback timing characteristics to construct a data-driven two-way feedback mechanism. This module first executes a user identity verification process. By integrating multi-source heterogeneous data (including community geographic location information, social group identifiers, and user nicknames, combined with real-time updated streaming message data), and based on supplementary user identity data in the database, it achieves accurate matching and verification of user identity, ensuring the security and relevance of subsequent responses. If user identity verification is successful, the system automatically collects key data such as the user's feedback timestamp and geographic coordinates as supplementary fields to the input data.

[0059] The data integration module accesses real-time environmental information (residential area environmental data) of the power consumption area through multi-source data interfaces. In response to extreme weather conditions such as high temperature and severe cold, as well as unplanned power outages and other emergencies, if the real-time environment is in extreme weather or unplanned power outages, it can be used as a contextualized supplementary value (scenario correction parameter) during subsequent data integration to form multi-dimensional data sentiment level labels.

[0060] The entire data integration module of the intelligent response system is based on the emotion level label data output in step 2. After using user identity supplementary data to determine the user's identity, it uses power consumption area environmental data and historical user feedback data to generate contextualized supplementary values ​​(scene correction parameters), and finally forms a multi-dimensional emotion level label (adjusted emotion level label data) that combines the characteristics of historical power outage accidents and real-time environmental attributes.

[0061] The data integration module in the intelligent response system achieves full-link coverage from basic information collection to scenario-based data fusion through dynamic data supplementation and contextualized supplementary value embedding. This ensures that the data processing process has both spatiotemporal accuracy and environmental adaptability, providing high-quality data support that is structured, scenario-based, and user-friendly for subsequent intelligent responses.

[0062] The feedback information generation module performs in-depth comparative analysis between the received processed fault feature itemsets (basic itemsets, combined itemsets, and context itemsets) and the historical power outage event database (also known as historical event data), covering multi-dimensional information such as event occurrence time, impact range, fault type, and handling records. The system can quickly locate similar historical events, extract fault handling experience and strategies, and provide solution references for current faults.

[0063] During the feedback generation phase, the system employs a tiered response strategy, dynamically adjusting the feedback content and priority based on the integrated multi-dimensional data emotion level labels. The newly generated emotion level labels (adjusted emotion level label data) are built upon a dynamic evaluation framework of multi-source data fusion. Their values ​​not only depend on the basic emotion level output by emotion recognition and emotion level calculation technologies but also require dynamic calibration through contextual supplementary rules generated by the data integration module.

[0064] Specifically, based on the association rule engine and historical case library (also known as historical event data), the system enhances the basic sentiment level for negative situational features such as extreme weather (such as high temperature and cold wave) and unplanned power outages that have not been resolved for a long time through a weight gain mechanism.

[0065] For positive contextual factors such as pleasant weather and short-term planned power outages, a weight decay strategy is used to lower the predicted sentiment level, forming a two-layer mapping mechanism of "textual sentiment features - supplementary environmental contextual features". This dynamic calibration process ensures that the sentiment level labels can comprehensively reflect the semantic connotation of the textual feedback and the actual scenario pressure, improving the environmental adaptability and relevance of the classification results.

[0066] Based on the adjusted sentiment level label data, the operation and maintenance strategy is set as follows: The feedback information generation module provides intelligent responses. If the user's emotional level is low, the system generates standardized and concise information responses, focusing on factual descriptions (such as the estimated recovery time and a brief description of the cause of the failure).

[0067] If the user's emotional level is high, the system generates a response containing empathetic expressions and prioritizes pushing it to the power supply department's dispatch system, while simultaneously triggering an expedited processing procedure. The urgent information fed back to the power supply department focuses more on technical details and handling suggestions, including fault location data, historical handling solutions, and the current user's emotional state, providing data support for operational and maintenance decisions.

[0068] This invention significantly reduces the complexity of natural language processing and response strategy generation by inputting concise and structured information, thus optimizing the decision-making path. Through differentiated response designs for user-side and power supply-side needs, it effectively improves fault handling efficiency and user satisfaction, achieving precise and efficient intelligent response in power service scenarios.

[0069] Compared with the prior art, the present invention has the following beneficial effects: 1. Enhance user responsiveness and system iteration. By introducing emotion level calculation and knowledge reasoning, user emotions can be accurately identified and their needs transformed, enabling intelligent scheduling of fault handling and personalized feedback. A feedback loop is established, collecting historical user feedback information and using this information to improve the diagnostic model, driving continuous system optimization, ensuring reasonable feedback generation, and significantly improving user experience and system adaptability.

[0070] 2. Build an integrated intelligent operation and maintenance ecosystem By integrating functions across multiple stages and based on precise diagnosis, personalized response, and closed-loop optimization, an intelligent operation and maintenance system covering the entire process is formed, effectively solving the problem of insufficient collaboration in existing technologies and promoting the development of power operation and maintenance towards a more intelligent and efficient direction.

[0071] 3. Construct a feedback response method that is both targeted and timely. By leveraging emotion recognition and emotion level calculation technologies, key features such as time and location words in user feedback are extracted. Combined with temporal feature analysis, this enables accurate determination of the timeliness of feedback information. A scenario-based response model is constructed based on environmental variables such as electricity usage time and user type. For high-priority and high-urgency user feedback, a decision-making mechanism integrating fault characteristics, emotion levels, and environmental factors generates differentiated intelligent responses. This forms a closed-loop processing flow of "feature extraction - timeliness determination - environmental adaptation - intelligent response" in the power grid operation and maintenance system, effectively improving the relevance and timeliness of feedback responses.

[0072] Example 5: Please see Figure 7As shown, the present invention also provides an electronic device 100 for a residential power supply anomaly maintenance method; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0073] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the residential power supply anomaly maintenance method described in Embodiment 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0074] The at least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor. The processor 102 is the control center of the electronic device 100, connecting various parts of the electronic device 100 via various interfaces and lines.

[0075] The memory 101 in the electronic device 100 stores multiple instructions to implement a method for abnormal operation and maintenance of residential power supply, and the processor 102 can execute the multiple instructions to achieve the following: Obtain information on power outages reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data; Emotion recognition is performed on power outage information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words; Based on fault location information and sentiment vocabulary, basic itemsets, combined itemsets, and contextual itemsets are constructed. Based on basic itemsets, combined itemsets, and contextual itemsets, frequent itemsets and strong association rules are determined. Frequent itemsets and strong association rules are mapped to sentiment level labels to obtain sentiment level label data. Based on real-time residential area environmental data and historical user feedback data, scene correction parameters are obtained; After using supplementary user identity data to determine the user's identity, the emotion level label data is adjusted based on the scenario correction parameters to obtain the adjusted emotion level label data. The adjusted emotion level label data and historical event data are compared to obtain the comparison results, and operation and maintenance strategies are set based on the comparison results.

[0076] Example 6: If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0077] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0078] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0079] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0080] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for maintaining and repairing abnormal power supply in residential areas, characterized in that, Includes the following steps: Obtain information on power outages reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data; Emotion recognition is performed on power outage information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words; Based on fault location information and sentiment vocabulary, basic itemsets, combined itemsets, and contextual itemsets are constructed. Based on basic itemsets, combined itemsets, and contextual itemsets, frequent itemsets and strong association rules are determined. Frequent itemsets and strong association rules are mapped to sentiment level labels to obtain sentiment level label data. Based on real-time residential area environmental data and historical user feedback data, scene correction parameters are obtained; After using supplementary user identity data to determine the user's identity, the emotion level label data is adjusted based on the scenario correction parameters to obtain the adjusted emotion level label data. The adjusted emotion level label data and historical event data are compared to obtain the comparison results, and operation and maintenance strategies are set based on the comparison results.

2. The method for maintaining abnormal power supply to residents according to claim 1, wherein the step of performing emotion recognition on the power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words specifically involves: using a BERT model to perform emotion recognition on the power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words.

3. The method for maintaining and repairing abnormal power supply in residential areas according to claim 1, characterized in that, Before constructing basic itemsets, combined itemsets, and contextual itemsets based on fault location information and sentiment lexicon, stop word filtering and part-of-speech normalization are performed on the fault location information and sentiment lexicon.

4. The method for maintaining and repairing abnormal power supply in residential areas according to claim 1, characterized in that, The determination of frequent itemsets and strong association rules based on basic itemsets, combined itemsets, and contextual itemsets specifically involves: From the basic itemset, combined itemset, and contextual itemset, filter out itemsets that are not frequent enough and retain only the frequently occurring itemsets to obtain frequent itemsets; Based on frequent itemsets, the confidence level is determined, and then the strong association rules are determined based on the confidence level.

5. The method for maintaining and repairing abnormal power supply in residential areas according to claim 1, characterized in that, The process of mapping frequent itemsets and strong association rules to sentiment level labels to obtain sentiment level label data specifically includes: Map frequent itemsets and strong association rules to predicted sentiment levels; Based on the predicted values ​​of the emotion levels, emotion level label data is obtained.

6. The method for maintaining and repairing abnormal power supply in residential areas according to claim 5, characterized in that, The formula for calculating the predicted value of mood level is: in, This is a predicted value for mood level. For the first i The weights of each feature, The result is the normalization of the features.

7. The method for maintaining abnormal power supply in residential areas according to claim 1, characterized in that, The adjusted emotion level label data is compared with historical event data to obtain comparison results. Based on the comparison results, an operation and maintenance strategy is set, specifically as follows: If the user's emotional level is low, provide a standardized and concise response. If the user's emotional level is high, respond with empathetic expressions and report to the dispatch department for expedited processing.

8. A residential power supply anomaly operation and maintenance system, characterized in that, include: The data acquisition module is used to acquire power outage fault information reported by residents, real-time environmental data of residential areas, supplementary user identity data, historical user feedback data, and historical event data; The emotion recognition module is used to perform emotion recognition on power outage fault information reported by residents and historical user feedback data to obtain fault location information and a set of emotional words. The itemset construction module is used to construct basic itemsets, composite itemsets, and contextual itemsets based on fault location information and sentiment vocabulary sets. The rule determination module is used to determine frequent itemsets and strong association rules based on basic itemsets, combined itemsets, and contextual itemsets. The mapping module is used to map frequent itemsets and strong association rules to sentiment level labels, thereby obtaining sentiment level label data. The scene correction parameter determination module is used to obtain scene correction parameters based on real-time residential area environmental data and user historical feedback data; The adjustment module is used to determine the user's identity using supplementary user identity data, and then adjust the emotion level label data based on the scenario correction parameters to obtain the adjusted emotion level label data. The operations and maintenance module is used to compare the adjusted emotion level label data with historical event data, obtain the comparison results, and set operations and maintenance strategies based on the comparison results.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the residential power supply anomaly operation and maintenance method as described in any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the residential power supply anomaly operation and maintenance method as described in any one of claims 1 to 7.

Images