Early warning method and system based on electric power engineering digital auditing and AI

Abstract

The invention discloses an early warning method and system based on electric power engineering digital auditing and AI. The method comprises the steps that current engineering data and historical engineering data of a target power grid are acquired; determining a current data abstract corresponding to the current engineering data; analyzing the current data abstract by using an electric power engineering digital auditing system to obtain auditing opinion information; determining a second target object from a plurality of first target objects of the target power grid based on the audit opinion information; based on the current engineering data, the historical engineering data, the auditing opinion information and the object information of the second target object, utilizing an artificial intelligence model to generate early warning prompt information and a candidate adjustment scheme; and engineering risk early warning is carried out based on the early warning prompt information and the candidate adjustment scheme, so that related audit can be automatically completed and a corresponding engineering risk early warning result can be given, accurate early warning can be carried out on risks existing in the electric power engineering, and the reliability of an engineering construction scheme is improved.

Description

Technical Field

[0001] The present application relates to the field of electric power technology, and in particular to an early warning method and system based on digital auditing and AI of electric power engineering. Background Art

[0002] At present, after the plan for a power engineering project is determined, manual audits are often relied on to review whether the plan meets the requirements (such as safety, design, construction, supervision, etc.). Manual audits require manual comparison of various requirements and standards (such as safety, design, construction, supervision, etc.) to check each parameter and operation process in the plan one by one. This method is extremely inefficient. Faced with the increasingly large and complex power engineering data, the manual processing speed is far behind the actual needs. At the same time, the accuracy of manual audits is greatly affected by factors such as the professional level and working status of auditors, and it is easy to make omissions, making it difficult to fully guarantee the reliability of power engineering audits.

[0003] With the development of AI (Artificial Intelligence) technology, although it has been partially applied in the field of engineering auditing, there are obvious deficiencies in its integration with digital auditing of power engineering projects. For example, it is difficult for existing AI algorithms to accurately adapt to the special rules of power engineering auditing, resulting in the inability to effectively play the advantages of AI in the process of engineering risk warning and solution generation, and unable to provide reliable and accurate support for the adjustment of power engineering-related project solutions. Summary of the invention

[0004] In order to solve the above technical problems, the embodiment of the present application proposes an early warning method and system based on digital auditing and AI of power engineering, which can accurately warn of the risks existing in power engineering and improve the reliability of engineering construction plans.

[0005] In a first aspect, the embodiment of the present application provides an early warning method based on digital auditing and AI of electric power engineering, including: Acquire current engineering data and historical engineering data of the target power grid, wherein the current engineering data is obtained by performing text recognition and semantic understanding on the data of the current engineering of the target power grid; Determining a current data summary corresponding to the current engineering data; Utilizing the electric power engineering digital audit system, analyzing the current data summary to obtain audit opinion information; Based on the audit opinion information, determining a second target object from a plurality of first target objects of the target power grid; Based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object, using an artificial intelligence model to generate early warning information and at least one candidate adjustment plan; Based on the early warning information and the at least one candidate adjustment plan, an engineering risk early warning is performed, wherein the early warning result is used to adjust the construction plan of the current project.

[0006] Optionally, the performing of engineering risk warning based on the warning prompt information and the at least one candidate adjustment scheme includes: According to the early warning prompt information, respectively simulate the at least one candidate adjustment scheme to obtain a simulation result of each of the at least one candidate adjustment scheme; Determining an adjustment plan based on the simulation result, the early warning prompt information and the at least one candidate adjustment plan; Conduct engineering risk warning according to the adjustment plan.

[0007] Optionally, the historical engineering data includes a plurality of historical sub-data, and the simulating the at least one candidate adjustment scheme respectively according to the early warning prompt information to obtain the simulation result of each of the at least one candidate adjustment scheme includes: For each of the candidate adjustment solutions, Determining at least one historical sub-data matching the candidate adjustment scheme from the plurality of historical sub-data; Determining at least one simulation model corresponding to the candidate adjustment scheme based on the at least one historical sub-data; According to the early warning prompt information, calling the at least one simulation model to simulate the candidate adjustment schemes respectively, and obtaining simulation data corresponding to each of the at least one simulation model; A simulation result of the candidate adjustment solution is determined according to the simulation data.

[0008] Optionally, determining an adjustment scheme based on the simulation result, the early warning prompt information and the at least one candidate adjustment scheme includes: For each of the candidate adjustment schemes, a difference analysis is performed between the simulation result and the early warning prompt information to obtain a difference analysis result, and the candidate adjustment scheme is verified according to the difference analysis result; Based on the at least one candidate adjustment solution and its respective verification result, an adjustment solution is determined.

[0009] Optionally, the generating early warning information and at least one candidate adjustment plan by using an artificial intelligence model based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object includes: generating key information based on the audit opinion information and the object information; Based on the key information and the historical engineering data, determining a target key scene from a plurality of preset key scenes; Based on the target key scenario, the current engineering data and the historical engineering data, the artificial intelligence model is used to generate the early warning prompt information and the at least one candidate adjustment plan.

[0010] Optionally, the determining a target key scene from a plurality of preset key scenes based on the key information and the historical engineering data includes: Determining at least one candidate key scene matching the key information from the multiple key scenes; Determining an evaluation value of each of the at least one candidate key scenes; Based on the historical engineering data and the evaluation value of each of the at least one candidate key scene, a target key scene is determined from the at least one candidate key scene.

[0011] Optionally, the artificial intelligence model includes a large model, and the generating of the early warning prompt information and the at least one candidate adjustment plan by using the artificial intelligence model based on the target key scenario, the current engineering data and the historical engineering data includes: The target key scenario, the current engineering data and the historical engineering data are input into the large model, so that the large model analyzes the current engineering data and the historical engineering data under the prompt of the target key scenario to generate the early warning prompt information and the at least one candidate adjustment plan.

[0012] Optionally, the using of the electric power engineering digital audit system to analyze the current data summary to obtain audit opinion information includes: Inputting the current data summary into the electric power engineering digital audit system; The electric power engineering digital audit system expands at least part of the information of the current data summary; The electric power engineering digital audit system performs a query based on the expanded information to obtain the audit opinion information.

[0013] Optionally, the electric power engineering digital audit system is configured as follows: Using the preset audit key node analysis strategy, extract information from the current data summary to obtain the original information sequence; Get audit prompt information; The original information sequence is expanded based on the audit prompt information to obtain the expanded information.

[0014] In a second aspect, the embodiment of the present application provides an early warning system based on digital auditing and AI of electric power engineering, including: A data acquisition module, used to acquire current engineering data and historical engineering data of a target power grid, wherein the current engineering data is obtained by performing text recognition and semantic understanding on the data of the current engineering of the target power grid; A data summary determination module, used to determine a current data summary corresponding to the current engineering data; An audit module, used to analyze the current data summary using a digital audit system for electric power engineering to obtain audit opinion information; A second target object determination module, configured to determine a second target object from a plurality of first target objects of the target power grid based on the audit opinion information; An intelligent analysis module, configured to generate early warning information and at least one candidate adjustment plan using an artificial intelligence model based on the current engineering data, the historical engineering data, the audit opinion information, and the acquired object information of the second target object; The early warning module is used to carry out engineering risk early warning based on the early warning prompt information and the at least one candidate adjustment plan, wherein the early warning result is used to adjust the construction plan of the current project.

[0015] In summary, the embodiments of the present application have at least the following beneficial effects: According to an embodiment of the present application, current engineering data and historical engineering data of a target power grid are acquired; a current data summary corresponding to the current engineering data is determined; a digital audit system for electric power engineering is used to analyze the current data summary to acquire audit opinion information; based on the audit opinion information, a second target object is determined from a plurality of first target objects of the target power grid; based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object, an artificial intelligence model is used to generate early warning information and at least one candidate adjustment plan; based on the early warning information and the at least one candidate adjustment plan, an engineering risk early warning is performed, wherein the early warning result is used to adjust the construction plan of the current project, so that relevant audits can be automatically completed and corresponding engineering risk early warning results can be given, so as to accurately warn of the risks existing in the electric power project and improve the reliability of the engineering construction plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart of an early warning method based on digital auditing and AI of electric power engineering provided in an embodiment of the present application; Figure 2 It is a structural diagram of an early warning system based on digital auditing and AI of electric power engineering provided in an embodiment of the present application; Figure 3 It is a schematic diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] In the description of the present application, the terms "first", "second", "third", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, features defined as "first", "second", "third", etc. may explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "multiple" is two or more. In the description of the present application, the term "including" and its variations are open inclusions, i.e., "including but not limited to". The term "based on" means "at least partially based on". The term "according to" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments".

[0019] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0020] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this application have the same meaning as those commonly understood by those skilled in the art. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by specific circumstances.

[0021] First, see Figure 1, shows a flow chart of an early warning method based on digital auditing of electric power engineering and AI (Artificial Intelligence) provided in an embodiment of the present application, the method includes steps S101-S106, which are as follows: S101, obtaining current engineering data and historical engineering data of a target power grid, wherein the current engineering data is obtained by performing text recognition and semantic understanding on the data of the current engineering of the target power grid.

[0022] In one example, the current project data may include at least one of the following: project establishment data, design data, construction data, supervision data, settlement data, etc. of the current project of the target power grid. Accordingly, the information of the current project may include at least one of the following: project establishment data, design data, construction data, supervision data, settlement data, etc. of the current project of the target power grid, wherein these data may be stored in the form of text and / or image, and the text recognition may include text recognition for images and / or text recognition for text (such as text keyword recognition).

[0023] In one example, the historical project data may include at least one of the following: project approval data, design data, construction data, supervision data, settlement data, and other data of historical projects of the target power grid.

[0024] S102, determining a current data summary corresponding to the current engineering data. The current data summary can remove non-critical information in the current engineering data, so that the subsequent power engineering digital audit system can perform more efficient and accurate audit analysis.

[0025] In one example, the current data summary may be information for describing key features of the current engineering data. In this case, the current data summary may be obtained by extracting features from the current engineering data using a common feature extraction method.

[0026] In one example, the method described in the present application can be executed by an electric power engineering scheme generation system. Since the electric power engineering digital audit system is usually deployed separately from the electric power engineering scheme generation system, it is necessary to ensure the security of data communication between the two systems. In this way, the determination of the current data digest corresponding to the current engineering data may include: determining the current data digest based on the current engineering data using a message digest algorithm to ensure the data integrity and non-tampering of the data corresponding to the current data digest, thereby ensuring that the electric power engineering digital audit system can receive a safe and reliable current data digest.

[0027] S103, using the electric power engineering digital audit system, analyzing the current data summary to obtain audit opinion information.

[0028] In one example, the above-mentioned electric power engineering digital audit system may be pre-configured with automatic audit software, so that the electric power engineering digital audit system can be configured to automatically audit and analyze the current data summary according to the set audit rules, so as to issue corresponding audit opinion information and feedback to the above-mentioned electric power engineering solution generation system, wherein the automatic audit software may be a relevant software program pre-written by engineering personnel according to the audit rules in the field of electric power engineering. The audit rules may be used to audit at least one of the project establishment data, design data, construction data, supervision data, and settlement data.

[0029] Project establishment data may include project feasibility study reports, project proposals, etc. Project establishment data can be used to evaluate the necessity and feasibility of the project.

[0030] Design data may include design plans, drawings, etc. of power engineering projects. Design data may be used to ensure that the design complies with regulatory requirements (such as safety requirements for the project) and technical standards.

[0031] Construction data may include various records and reports during the construction process, such as construction logs, quality inspection records, etc. Construction data may be used to indicate construction quality and progress.

[0032] Settlement data may include financial settlement information related to the project.

[0033] In one example, the automatic audit software can also be obtained by fine-tuning a general large language model using the collected knowledge in the field of power engineering auditing. The fine-tuning process is a common way of fine-tuning a large language model in this field.

[0034] S104, based on the audit opinion information, determining a second target object from a plurality of first target objects of the target power grid; wherein the first target object may include various types of power generation equipment to be adjusted.

[0035] In one example, the audit opinion information can be used to indicate the first target object (i.e., the second target object) that needs to be focused on among the multiple first target objects confirmed by the audit, among the relevant adjustment plans for the current project corresponding to the current project data. For example, in the construction project, if the audit finds that there are problems with the installation of high-voltage equipment (such as transformers, circuit breakers, etc.), then the high-voltage equipment needs to be taken as the second target object for key adjustment and attention, and / or, if the audit finds that the cable and wiring system (such as cable selection, laying path and joint processing) has cable specifications that do not meet the design requirements and / or the wiring has potential safety hazards, then the cable and wiring system will be taken as the second target object, and / or, if the audit finds that the grounding resistance value in the grounding and lightning protection system exceeds the standard or the lightning protection device is improperly installed, then the grounding and lightning protection system will be taken as the second target object.

[0036] S105, based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object, using an artificial intelligence model to generate early warning information and at least one candidate adjustment plan.

[0037] In one example, the artificial intelligence model can be pre-trained via sample data and labels, so that the trained artificial intelligence model can take current engineering data, historical engineering data, audit opinion information and object information as input, and output early warning information and at least one candidate adjustment plan in response to the input.

[0038] Among them, the sample data may include sample current engineering data, sample historical engineering data, sample audit opinion information and sample object information, and the labels include sample early warning prompt information and sample adjustment plans. The specific training process may include: inputting the sample data into the artificial intelligence model to be trained to obtain the output of the artificial intelligence model; based on the difference between the output of the artificial intelligence model and the above-mentioned labels, using the minimization loss function to optimize the model parameters of the artificial intelligence model to be trained, thereby completing the model training.

[0039] S106, based on the early warning information and the at least one candidate adjustment plan, perform an engineering risk early warning, wherein the early warning result is used to adjust the construction plan of the current project.

[0040] In one example, the early warning results here can be used to characterize whether the construction safety of the current project (such as the construction safety related to the above-mentioned high-voltage equipment, cables and wiring systems, grounding and lightning protection systems) can be ensured when applying various candidate adjustment plans, so as to facilitate adjustments to unsafe areas in the construction plan.

[0041] In one example, based on the early warning information and the at least one candidate adjustment plan, performing an engineering risk early warning may include: selecting a candidate adjustment plan that matches the early warning information from the at least one candidate adjustment plan, and performing an engineering risk early warning according to the selected candidate adjustment plan, wherein the early warning result is used to adjust the construction plan of the current project. The candidate adjustment plan that matches the early warning information may refer to the candidate adjustment plan with the highest similarity to the early warning information.

[0042] In an optional implementation, the performing of engineering risk warning based on the warning prompt information and the at least one candidate adjustment plan includes: According to the early warning prompt information, the at least one candidate adjustment scheme is simulated respectively to obtain a simulation result of each of the at least one candidate adjustment scheme.

[0043] An adjustment scheme is determined based on the simulation result, the early warning prompt information and the at least one candidate adjustment scheme.

[0044] Conduct engineering risk warning according to the adjustment plan.

[0045] In one example, according to the early warning information, simulating the at least one candidate adjustment scheme respectively to obtain the simulation result of each of the at least one candidate adjustment scheme may include: setting the simulation model according to the early warning information, and using the set simulation model to simulate each candidate adjustment scheme respectively to obtain the corresponding simulation result. Exemplarily, the above-mentioned simulation model may include general engineering construction simulation software.

[0046] In an optional implementation, the historical engineering data includes a plurality of historical sub-data, and the simulating the at least one candidate adjustment scheme according to the early warning prompt information to obtain the simulation result of each of the at least one candidate adjustment scheme includes: For each of the candidate adjustment solutions, At least one historical sub-data matching the candidate adjustment scheme is determined from the plurality of historical sub-data.

[0047] At least one simulation model corresponding to the candidate adjustment scheme is determined based on the at least one historical sub-data.

[0048] According to the early warning information, the at least one simulation model is called to simulate the candidate adjustment schemes respectively to obtain simulation data corresponding to each of the at least one simulation model.

[0049] A simulation result of the candidate adjustment solution is determined according to the simulation data.

[0050] In one example, the at least one simulation model corresponds one-to-one to at least one configuration parameter, and the at least one configuration parameter corresponds one-to-one to at least one historical sub-data, wherein each configuration parameter is determined by the corresponding historical sub-data and is used to configure the corresponding simulation model. In addition, at least one simulation model can be set according to the early warning information, and the set simulation model can be used to simulate the candidate adjustment scheme to obtain the corresponding simulation data. In this way, the simulation model used for simulation can be configured with parameters corresponding to the early warning information and the historical sub-data to complete the simulation as accurately as possible.

[0051] In one example, the multiple historical sub-data included in the historical engineering data can be obtained by dividing the historical engineering data according to time and / or engineering projects. In this way, each historical sub-data can represent the engineering data of the target power grid in a short period of time in the past or in a certain engineering project. At this time, the historical sub-data selected to match the candidate adjustment scheme represents a situation where the historical sub-data can be adapted to the candidate adjustment scheme in terms of certain characteristics.

[0052] In an optional implementation, determining an adjustment scheme based on the simulation result, the early warning prompt information and the at least one candidate adjustment scheme includes: For each of the candidate adjustment schemes, a difference analysis is performed between the simulation result and the early warning prompt information to obtain a difference analysis result, and the candidate adjustment scheme is verified based on the difference analysis result.

[0053] Based on the at least one candidate adjustment solution and its respective verification result, an adjustment solution is determined.

[0054] In one example, the above difference analysis results can be used to characterize the similarity between the normative warning risk points of each stage indicated by the calculated warning prompt information (for example, the risk points can be characterized in the form of portraits) and the simulation results, wherein the higher the similarity, the lower the value corresponding to the difference analysis result. In this way, the candidate adjustment plan that has passed the verification means that its normative warning risk points at each stage meet the requirements.

[0055] In one example, verifying the candidate adjustment scheme according to the difference analysis result may include: determining the candidate adjustment scheme whose difference analysis result satisfies the difference condition as the verified candidate adjustment scheme. The difference condition may include that the value corresponding to the difference analysis result is higher than the difference threshold, or, if the difference analysis result is a numerical value representing the degree of deviation, then a maximum allowable difference score may be specified, and the candidate adjustment scheme below the difference score is considered to be verified.

[0056] In one example, determining an adjustment scheme based on the at least one candidate adjustment scheme and its respective verification results may include: performing a safety assessment process on construction safety (e.g., construction safety related to the above-mentioned high-voltage equipment, cables and wiring systems, grounding and lightning protection systems) on each verified candidate adjustment scheme among the at least one candidate adjustment scheme to obtain a safety assessment value for each verified candidate adjustment scheme, and determining the candidate adjustment scheme with the highest safety assessment value as the adjustment scheme.

[0057] In an optional implementation, the generating of early warning information and at least one candidate adjustment plan using an artificial intelligence model based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object includes: Based on the audit opinion information and the object information, key information is generated.

[0058] Based on the key information and the historical engineering data, a target key scene is determined from a plurality of preset key scenes.

[0059] Based on the target key scenario, the current engineering data and the historical engineering data, the artificial intelligence model is used to generate the early warning prompt information and the at least one candidate adjustment plan.

[0060] In one example, the key scenario may be a preset typical scenario related to power engineering audit, such as "high-voltage equipment construction scenario", "cable and wiring system construction scenario", "grounding and lightning protection system construction scenario", etc.

[0061] High-voltage equipment construction scenarios may include at least one of the following: high-voltage equipment compliance inspection scenarios (checking whether the high-voltage equipment is correctly installed in accordance with technical specifications and standards), high-voltage equipment safety assessment scenarios (checking whether the safety and stability of equipment operation are assessed through strict testing and debugging records).

[0062] Cable and wiring system construction scenarios may include at least one of the following: specification inspection scenario (used to check whether the cable selection is consistent with the design requirements, to ensure that its current carrying capacity and insulation performance meet the actual use needs), safety hazard investigation scenario (identifying possible electrical fire risks or other safety hazards by reviewing the laying route diagram and checking the joint processing method), quality assessment scenario (performing insulation resistance test simulation to ensure that the cable maintains a good insulation state before and after installation, thereby improving the safety and reliability of the entire system).

[0063] The grounding and lightning protection system construction scenarios may include at least one of the following: grounding effect evaluation scenario (simulation based on the grounding resistance value to verify the effectiveness of the grounding system), lightning protection effectiveness inspection scenario (checking the installation location, quantity and protection range of the lightning protection device to ensure that buildings and electrical equipment are protected from lightning damage).

[0064] In this embodiment, key scenarios can be used to assist the artificial intelligence model in generating early warning information and candidate adjustment plans. By determining the target key scenarios, the model can focus on the specific situation corresponding to the current project, and combine the current project data with historical project data to more accurately analyze and generate early warning information and candidate adjustment plans that conform to the actual situation.

[0065] In one example, the key information can be obtained by integrating the audit opinion information and the object information.

[0066] In one example, based on the key information and the historical engineering data, determining the target key scene from a plurality of preset key scenes may include: calculating a first similarity between the key information and the information corresponding to each of the plurality of key scenes, and calculating a second similarity between the historical engineering data and the information corresponding to each of the plurality of key scenes, weighting the first similarity and the second similarity to obtain a third similarity, and determining the key scene with the highest third similarity as the target key scene. The key information and the historical engineering data each represent different types of information, and therefore, when calculating the similarity between the information corresponding to the key scene, the information similarity calculation may be performed for different information types.

[0067] In one example, the above-mentioned artificial intelligence model can be pre-trained through sample data and labels, so that the trained artificial intelligence model can take target key scenarios, current engineering data and historical engineering data as input, and output early warning prompt information and at least one candidate adjustment plan in response to the input.

[0068] The sample data of this embodiment may include sample current engineering data, sample historical engineering data and sample key scenarios, and the labels include sample warning prompt information and sample adjustment plans. The specific training process may include: inputting the sample data into the artificial intelligence model to be trained to obtain the output of the artificial intelligence model. Based on the difference between the output of the artificial intelligence model and the above labels, the model parameters of the artificial intelligence model to be trained are optimized by minimizing the loss function, thereby completing the model training.

[0069] In an optional implementation, the determining a target key scene from a plurality of preset key scenes based on the key information and the historical engineering data includes: At least one candidate key scene matching the key information is determined from the multiple key scenes.

[0070] An evaluation value of each of the at least one candidate key scenes is determined.

[0071] Based on the historical engineering data and the evaluation value of each of the at least one candidate key scene, a target key scene is determined from the at least one candidate key scene.

[0072] In one example, the evaluation value in this embodiment may include at least one of the following: a safety evaluation value, a benefit evaluation value, etc. The evaluation value may be obtained by evaluating at least one candidate key scenario using a corresponding evaluation model. The evaluation value may also be obtained by weighted calculation of the safety evaluation value and the benefit evaluation value.

[0073] In one example, based on the historical engineering data and the evaluation values ​​of the at least one candidate key scene, determining the target key scene from the at least one candidate key scene may include: determining each candidate key scene whose evaluation value is higher than the evaluation value threshold from the at least one candidate key scene, and inputting each candidate key scene whose evaluation value is higher than the evaluation value threshold and the historical engineering data into the large language model, so that the large language model analyzes each input candidate key scene under the prompt of the historical engineering data, so as to select the target key scene from each input candidate key scene according to the analysis result. The analysis result can be used to indicate safety and / or benefit, etc.

[0074] In an optional implementation, the artificial intelligence model includes a large model, and the generation of the early warning information and the at least one candidate adjustment plan using the artificial intelligence model based on the target key scenario, the current engineering data, and the historical engineering data includes: The target key scenario, the current engineering data and the historical engineering data are input into the large model, so that the large model analyzes the current engineering data and the historical engineering data under the prompt of the target key scenario to generate the early warning prompt information and the at least one candidate adjustment plan.

[0075] In one example, the above-mentioned big model may include a general big language model, and the target key scenarios may be converted into semantic information so as to serve as prompt words that can be understood by the big model. The prompt words can be used to prompt the big model to analyze current engineering data and historical engineering data, and make the analyzed results include early warning information and at least one candidate adjustment plan.

[0076] In an optional implementation manner, the using of the electric power engineering digital audit system to analyze the current data summary to obtain audit opinion information includes: The current data summary is input into the electric power engineering digital audit system.

[0077] The electric power engineering digital audit system expands at least part of the information of the current data summary.

[0078] The electric power engineering digital audit system performs a query based on the expanded information to obtain the audit opinion information.

[0079] In an optional implementation, the electric power engineering digital audit system is configured as follows: Using the preset audit key node analysis strategy, information is extracted from the current data summary to obtain the original information sequence.

[0080] Get audit prompt information.

[0081] The original information sequence is expanded based on the audit prompt information to obtain the expanded information.

[0082] In one example, expanding the original information sequence based on the audit prompt information to obtain the expanded information may include: performing content identification on the original information sequence to generate a content text; performing error analysis on the generated content text according to the standard audit text indicated by the audit prompt information; correcting the original information sequence based on the error analysis results; and expanding the corrected original information sequence to obtain the expanded information.

[0083] In one example, the specific process of generating audit opinion information may include the following aspects.

[0084] Information extraction: Use the preset audit key node analysis strategy to extract information from the current data summary and obtain the original information sequence. The audit key node analysis strategy here can be pre-set based on professional knowledge and experience in the field of power engineering, such as focusing on the key nodes in the current data summary related to the above-mentioned high-voltage equipment, cables and wiring systems, grounding and lightning protection systems.

[0085] Information expansion: obtain audit prompt information, and expand the original information sequence based on the audit prompt information. Specifically, it includes content recognition of the original information sequence to generate content text; error analysis of the generated content text according to the standard audit text indicated by the audit prompt information; based on the error analysis results, the original information sequence is corrected; the corrected original information sequence is expanded to obtain the expanded information. Audit prompt information can be used to indicate equipment operation safety standards, technical specifications, etc. in the field of power engineering.

[0086] On the second aspect, accordingly, the embodiment of the present application also provides an early warning system based on digital auditing and AI of power engineering, which can implement all processes of the early warning method based on digital auditing and AI of power engineering provided in the above embodiment.

[0087] See also Figure 2 , shows a schematic diagram of the structure of an early warning system based on digital audit and AI for electric power engineering provided in an embodiment of the present application, the early warning system based on digital audit and AI for electric power engineering including: The data acquisition module 201 is used to acquire the current engineering data and historical engineering data of the target power grid, wherein the current engineering data is obtained by performing text recognition and semantic understanding on the data of the current engineering of the target power grid.

[0088] The data summary determination module 202 is used to determine the current data summary corresponding to the current engineering data.

[0089] The audit module 203 is used to analyze the current data summary using the electric power engineering digital audit system to obtain audit opinion information.

[0090] The second target object determination module 204 is configured to determine a second target object from a plurality of first target objects of the target power grid based on the audit opinion information.

[0091] The intelligent analysis module 205 is used to generate early warning information and at least one candidate adjustment plan based on the current engineering data, the historical engineering data, the audit opinion information and the object information of the second target object obtained by using an artificial intelligence model.

[0092] The early warning module 206 is used to perform an early warning of engineering risks based on the early warning prompt information and the at least one candidate adjustment plan, wherein the early warning result is used to adjust the construction plan of the current project.

[0093] In an optional implementation, the performing of engineering risk warning based on the warning prompt information and the at least one candidate adjustment plan includes: According to the early warning prompt information, the at least one candidate adjustment scheme is simulated respectively to obtain a simulation result of each of the at least one candidate adjustment scheme.

[0094] An adjustment scheme is determined based on the simulation result, the early warning prompt information and the at least one candidate adjustment scheme.

[0095] Conduct engineering risk warning according to the adjustment plan.

[0096] In an optional implementation, the historical engineering data includes a plurality of historical sub-data, and the simulating the at least one candidate adjustment scheme according to the early warning prompt information to obtain the simulation result of each of the at least one candidate adjustment scheme includes: For each of the candidate adjustment solutions, At least one historical sub-data matching the candidate adjustment scheme is determined from the plurality of historical sub-data.

[0097] At least one simulation model corresponding to the candidate adjustment scheme is determined based on the at least one historical sub-data.

[0098] According to the early warning information, the at least one simulation model is called to simulate the candidate adjustment schemes respectively to obtain simulation data corresponding to each of the at least one simulation model.

[0099] A simulation result of the candidate adjustment solution is determined according to the simulation data.

[0100] In an optional implementation, determining an adjustment scheme based on the simulation result, the early warning prompt information and the at least one candidate adjustment scheme includes: For each of the candidate adjustment schemes, a difference analysis is performed between the simulation result and the early warning prompt information to obtain a difference analysis result, and the candidate adjustment scheme is verified based on the difference analysis result.

[0101] Based on the at least one candidate adjustment solution and its respective verification result, an adjustment solution is determined.

[0102] In an optional implementation, the generating of early warning information and at least one candidate adjustment plan using an artificial intelligence model based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object includes: Based on the audit opinion information and the object information, key information is generated.

[0103] Based on the key information and the historical engineering data, a target key scene is determined from a plurality of preset key scenes.

[0104] Based on the target key scenario, the current engineering data and the historical engineering data, the artificial intelligence model is used to generate the early warning prompt information and the at least one candidate adjustment plan.

[0105] In an optional implementation, the determining a target key scene from a plurality of preset key scenes based on the key information and the historical engineering data includes: At least one candidate key scene matching the key information is determined from the multiple key scenes.

[0106] An evaluation value of each of the at least one candidate key scenes is determined.

[0107] Based on the historical engineering data and the evaluation value of each of the at least one candidate key scene, a target key scene is determined from the at least one candidate key scene.

[0108] In an optional implementation, the artificial intelligence model includes a large model, and the generation of the early warning information and the at least one candidate adjustment plan using the artificial intelligence model based on the target key scenario, the current engineering data, and the historical engineering data includes: The target key scenario, the current engineering data and the historical engineering data are input into the large model, so that the large model analyzes the current engineering data and the historical engineering data under the prompt of the target key scenario to generate the early warning prompt information and the at least one candidate adjustment plan.

[0109] In one example, the artificial intelligence model includes a large model, for example, the large model adopts a common deep neural network structure, such as a multi-layer perceptron. Taking a simple multi-layer perceptron as an example, its specific structure may include an input layer, several hidden layers and an output layer.

[0110] Input layer: The number of nodes is determined by the dimension of the input data. The input data includes target key scenarios, current engineering data, and historical engineering data. Assuming that the feature dimension of these data after merging is n, the number of input layer nodes is n.

[0111] Hidden layer: 2-3 hidden layers can be set, and the number of nodes in each hidden layer can be adjusted based on experience and experiments, for example, set to 128 and 64 respectively. The function of the hidden layer is to extract features and perform nonlinear transformation on the input data.

[0112] Output layer: Output warning information and at least one candidate adjustment plan. Assuming that the warning information is represented by m_1 values, the candidate adjustment plan is represented by m_2 values, and k candidate adjustment plans are generated, the number of nodes in the output layer is m_1+k×m_2.

[0113] Correspondingly, the training algorithm of the multilayer perceptron can be divided into an optimizer part and a loss function part.

[0114] Optimizer: An optimizer that uses stochastic gradient descent. Stochastic gradient descent is a commonly used optimization algorithm that calculates the gradient by randomly selecting a portion of samples in each iteration and updating the model parameters according to the gradient, thereby gradually reducing the value of the loss function. Its update formula is ,in are the parameters of the model, is the learning rate, is the loss function with respect to the parameters gradient.

[0115] Loss function: Use mean square error loss function. For sample , whose true label is The model predicts that , then the mean square error loss function The calculation formula is ,in is the number of samples. By minimizing the mean square error loss function, the model's predicted value can be made as close to the true label as possible.

[0116] In an optional implementation manner, the using of the electric power engineering digital audit system to analyze the current data summary to obtain audit opinion information includes: The current data summary is input into the electric power engineering digital audit system.

[0117] The electric power engineering digital audit system expands at least part of the information of the current data summary.

[0118] The electric power engineering digital audit system performs a query based on the expanded information to obtain the audit opinion information.

[0119] In an optional implementation, the electric power engineering digital audit system is configured as follows: Using the preset audit key node analysis strategy, information is extracted from the current data summary to obtain the original information sequence.

[0120] Get audit prompt information.

[0121] The original information sequence is expanded based on the audit prompt information to obtain the expanded information.

[0122] In a third aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above-mentioned early warning methods based on digital auditing and AI of power engineering projects.

[0123] In a fourth aspect, an embodiment of the present application provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of any of the above-mentioned early warning methods based on digital auditing and AI of power engineering.

[0124] In a fifth aspect, an embodiment of the present application provides a computer device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the steps of any of the above-mentioned early warning methods based on digital auditing and AI of power engineering when executing the computer program.

[0125] See also Figure 3 The computer device of this embodiment includes: a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301, such as an early warning program based on digital auditing and AI for electric power engineering. When the processor 301 executes the computer program, the steps in the above-mentioned early warning method embodiments based on digital auditing and AI for electric power engineering are implemented, such as Figure 1 Steps S101-S106 are shown.

[0126] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory 302 and executed by the processor 301 to complete the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the computer device.

[0127] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will appreciate that the schematic diagram is merely an example of a computer device and does not constitute a limitation on the computer device. The computer device may include more or fewer components than shown in the diagram, or may combine certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0128] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor 301 may also be any conventional processor, etc. The processor 301 is the control center of the computer device, and uses various interfaces and lines to connect various parts of the entire computer device.

[0129] The memory 302 can be used to store the computer program and / or module. The processor 301 implements various functions of the computer device by running or executing the computer program and / or module stored in the memory 302 and calling the data stored in the memory 302. The memory 302 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.). In addition, the memory 302 can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0130] Wherein, if the module / unit integrated in the computer device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor 301. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0131] In summary, the embodiments of the present application have at least the following beneficial effects: According to an embodiment of the present application, current engineering data and historical engineering data of a target power grid are acquired; a current data summary corresponding to the current engineering data is determined; a digital audit system for electric power engineering is used to analyze the current data summary to acquire audit opinion information; based on the audit opinion information, a second target object is determined from a plurality of first target objects of the target power grid; based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object, an artificial intelligence model is used to generate early warning information and at least one candidate adjustment plan; based on the early warning information and the at least one candidate adjustment plan, an engineering risk early warning is performed, wherein the early warning result is used to adjust the construction plan of the current project, so that relevant audits can be automatically completed and corresponding engineering risk early warning results can be given, so as to accurately warn of the risks existing in the electric power project and improve the reliability of the engineering construction plan.

[0132] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary hardware platform, and of course it can also be implemented entirely by hardware. Based on such an understanding, all or part of the contribution of the technical solution of the present application to the background technology can be embodied in the form of a software product, and the computer software product can be stored in a storage medium, such as ROM (Read-Only Memory) / RAM (Random Access Memory), a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application or some parts of the embodiments.

[0133] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications are also considered to be within the scope of protection of the present application.

Claims

1. An early warning method based on digital auditing and AI of electric power engineering, characterized in that: include: Acquire current engineering data and historical engineering data of the target power grid, wherein the current engineering data is obtained by performing text recognition and semantic understanding on the data of the current engineering of the target power grid; Determining a current data summary corresponding to the current engineering data; Utilizing the electric power engineering digital audit system, analyzing the current data summary to obtain audit opinion information; Based on the audit opinion information, determining a second target object from a plurality of first target objects of the target power grid; Based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object, using an artificial intelligence model to generate early warning information and at least one candidate adjustment plan; Based on the early warning information and the at least one candidate adjustment plan, an engineering risk early warning is performed, wherein the early warning result is used to adjust the construction plan of the current project.

2. The method according to claim 1, characterized in that The performing of engineering risk warning based on the warning prompt information and the at least one candidate adjustment plan includes: According to the early warning prompt information, respectively simulate the at least one candidate adjustment scheme to obtain a simulation result of each of the at least one candidate adjustment scheme; Determining an adjustment plan based on the simulation result, the early warning prompt information and the at least one candidate adjustment plan; Conduct engineering risk warning according to the adjustment plan.

3. The method according to claim 2, characterized in that The historical engineering data includes a plurality of historical sub-data, and the at least one candidate adjustment scheme is simulated according to the early warning prompt information to obtain a simulation result of each of the at least one candidate adjustment scheme, including: For each of the candidate adjustment solutions, Determining at least one historical sub-data matching the candidate adjustment scheme from the plurality of historical sub-data; Based on the at least one historical sub-data, determining at least one simulation model corresponding to the candidate adjustment scheme; According to the early warning prompt information, calling the at least one simulation model to simulate the candidate adjustment schemes respectively, and obtaining simulation data corresponding to each of the at least one simulation model; A simulation result of the candidate adjustment solution is determined according to the simulation data.

4. The method according to claim 2, characterized in that: The determining of the adjustment scheme based on the simulation result, the early warning prompt information and the at least one candidate adjustment scheme includes: For each of the candidate adjustment schemes, a difference analysis is performed between the simulation result and the early warning prompt information to obtain a difference analysis result, and the candidate adjustment scheme is verified according to the difference analysis result; Based on the at least one candidate adjustment solution and its respective verification result, an adjustment solution is determined.

5. The method according to claim 1, characterized in that The generating of early warning information and at least one candidate adjustment plan by using an artificial intelligence model based on the current engineering data, the historical engineering data, the audit opinion information and the acquired object information of the second target object includes: generating key information based on the audit opinion information and the object information; Based on the key information and the historical engineering data, determining a target key scene from a plurality of preset key scenes; Based on the target key scenario, the current engineering data and the historical engineering data, the artificial intelligence model is used to generate the early warning prompt information and the at least one candidate adjustment plan.

6. The method according to claim 5, characterized in that The determining of a target key scene from a plurality of preset key scenes based on the key information and the historical engineering data includes: Determining at least one candidate key scene matching the key information from the multiple key scenes; Determining an evaluation value of each of the at least one candidate key scenes; Based on the historical engineering data and the evaluation value of each of the at least one candidate key scene, a target key scene is determined from the at least one candidate key scene.

7. The method according to claim 5, characterized in that The artificial intelligence model includes a large model, and based on the target key scenario, the current engineering data and the historical engineering data, the artificial intelligence model is used to generate the early warning prompt information and the at least one candidate adjustment plan, including: The target key scenario, the current engineering data and the historical engineering data are input into the large model, so that the large model analyzes the current engineering data and the historical engineering data under the prompt of the target key scenario to generate the early warning prompt information and the at least one candidate adjustment plan.

8. The method according to claim 1, characterized in that The method of using the electric power engineering digital audit system to analyze the current data summary to obtain audit opinion information includes: Inputting the current data summary into the electric power engineering digital audit system; The electric power engineering digital audit system expands at least part of the information of the current data summary; The electric power engineering digital audit system performs a query based on the expanded information to obtain the audit opinion information.

9. The method according to claim 8, characterized in that The electric power engineering digital audit system is configured to: Using the preset audit key node analysis strategy, extract information from the current data summary to obtain the original information sequence; Get audit prompt information; The original information sequence is expanded based on the audit prompt information to obtain the expanded information.

10. An early warning system based on digital auditing and AI for electric power engineering, characterized in that: include: A data acquisition module, used to acquire current engineering data and historical engineering data of a target power grid, wherein the current engineering data is obtained by performing text recognition and semantic understanding on the data of the current engineering of the target power grid; A data summary determination module, used to determine a current data summary corresponding to the current engineering data; An audit module, used to analyze the current data summary using a digital audit system for electric power engineering to obtain audit opinion information; A second target object determination module, configured to determine a second target object from a plurality of first target objects of the target power grid based on the audit opinion information; An intelligent analysis module, configured to generate early warning information and at least one candidate adjustment plan using an artificial intelligence model based on the current engineering data, the historical engineering data, the audit opinion information, and the acquired object information of the second target object; The early warning module is used to carry out engineering risk early warning based on the early warning prompt information and the at least one candidate adjustment plan, wherein the early warning result is used to adjust the construction plan of the current project.

Images