Operation risk management and control method and device, computer equipment, medium and program

Abstract

The invention relates to the technical field of power grid operation risk management and control, and discloses an operation risk management and control method and device, computer equipment, a medium and a program. The method is used for evaluating the risk in the construction process of the power distribution network equipment. The method comprises the following steps: acquiring target construction data generated in the construction process of the power distribution network equipment; extracting the embedding of the target construction data by using the multi-modal embedding model to obtain a construction embedding; inputting the construction embedding into a risk assessment model, and determining a target result of risk assessment through the risk assessment model; the risk assessment model is obtained by training historical data, the historical data comprises embedding of historical construction data generated in the construction process of the power distribution network equipment and a historical assessment result, and the historical assessment result is a result obtained by risk assessment based on the historical construction data. Based on the technical scheme of the invention, the risk assessment efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of power grid operation risk management and control, and specifically to operation risk management and control methods, devices, computer equipment, media and programs. Background Art

[0002] With the progress and development of social economy, the electricity load continues to grow, and the number of new construction, expansion, technical transformation and overhaul construction continues to increase. For a long time, due to the large number of distribution network equipment, the high complexity of construction, heavy tasks, short time, and many dangerous points, the risk of casualties and electrical equipment failures remains high, which brings great difficulties to the safety management of on-site construction operations. How to ensure the safe development of construction operations is the primary issue facing safe production. In order to do a good job in on-site operation risk control, it is necessary to start from two aspects. On the one hand, we need to continuously strengthen the technical level assessment of operators and the competence assessment of managers; on the other hand, we need to standardize and strengthen the control of on-site operations. How to assess the risks of on-site operations and how to carry out differentiated control according to different risk levels are the biggest challenges faced by managers. In this context, the assessment model of on-site operation risks becomes the key core point.

[0003] At present, the risk assessment process relies too much on the construction data generated during the construction of distribution network equipment. However, due to the large number of data modes in the construction data, this brings great difficulties to automated risk assessment.

[0004] Therefore, there is an urgent need for an operational risk management method that can improve the level of risk assessment automation. Summary of the invention

[0005] Therefore, the technical problem to be solved by the present application is to overcome the problems of low intelligence level and low accuracy in operation risk assessment in related technologies.

[0006] In order to solve the above technical problems, in a first aspect, the present application provides an operation risk management method, which is used to evaluate the risks in the process of constructing distribution network equipment. The operation risk management method includes:

[0007] Acquire target construction data generated during the construction of the distribution network equipment; the target construction data is data including more than one mode, the data type of the mode characterizing the data;

[0008] Extracting the embedding of the target construction data using a multimodal embedding model to obtain a construction embedding; the multimodal embedding model is used to extract embedding of data of more than one modality;

[0009] The construction is embedded into a risk assessment model, and the target result of the risk assessment is determined through the risk assessment model; the risk assessment model is trained using historical data, and the historical data includes the embedding of historical construction data generated during the construction of the distribution network equipment and historical assessment results, and the historical assessment results are the results of risk assessment based on the historical construction data.

[0010] In combination with any implementation manner of the present application, the acquiring of target construction data generated during the construction of the distribution network equipment includes:

[0011] The target construction data generated during the construction of the distribution network equipment is collected by using a collection device; the target construction data includes personnel information data, construction equipment data, and construction environment data, the personnel information data includes: information of construction personnel who construct the distribution network equipment, the construction equipment data includes: data generated by the distribution network equipment during the construction of the distribution network equipment; the construction environment data represents data of the environment during the construction of the distribution network equipment, the collection device includes a handheld terminal and a communication terminal, the handheld terminal is used to collect and store the personnel information data, and the communication terminal is used to collect and store the construction equipment data and the construction environment data.

[0012] In combination with any implementation manner of the present application, the step of extracting the embedding of the target construction data using a multimodal embedding model to obtain the construction embedding includes:

[0013] extracting embeddings of the target construction data using the multimodal embedding model to obtain candidate embeddings;

[0014] The candidate embedding is divided into n groups to obtain n reference embeddings; the n groups of reference embeddings include a first embedding, the first embedding includes m first vectors, and n is a positive integer greater than 1;

[0015] Determine j second vectors for representing m first vectors from the first embedding; j is greater than or equal to 1, and m is a positive integer greater than j and greater than 1;

[0016] The construction embedding is determined based on j second vectors and third vectors; the third vector is used to represent a second embedding, and the second embedding is an embedding other than the first embedding in the n groups of reference embeddings.

[0017] In conjunction with any embodiment of the present application, the target result includes a target risk level of the construction risk of the current construction determined by a risk assessment model, and the method further includes:

[0018] In the case where the target risk level does not meet the preset warning level, the construction embedding is input into a trend prediction model, and the embedding change trend of the construction embedding is predicted by the trend prediction model; the embedding change trend is the change trend of the construction embedding after the current moment;

[0019] Determine, according to the embedded change trend, a prediction result of the construction risk generated by the current construction after the current moment;

[0020] The prediction results are sent to construction personnel who are constructing the distribution network equipment, so that the construction personnel can check the construction risks according to the prediction results during the construction process.

[0021] In combination with any embodiment of the present application, the method further includes:

[0022] When the target risk level meets the preset warning level, the target result is sent to the construction personnel who are constructing the distribution network equipment, so that the construction personnel can rectify the construction risk according to the target result;

[0023] Obtaining rectification data generated by the construction personnel in the process of rectifying the construction risks according to the target results;

[0024] extracting an embedding of the rectification data using the multimodal embedding model to obtain a rectification embedding;

[0025] The rectification is embedded into the risk assessment model, and the rectification result of the risk assessment is determined by the risk assessment model.

[0026] In combination with any embodiment of the present application, the modal embedding model includes a first embedding model, a second embedding model and a fusion model, the first embedding model is used to extract the embedding of first data, the first data is data of a first data type, the second embedding model is used to extract the embedding of second data, the second data is data of a second data type, and the fusion model is used to fuse the embedding of the first data and the embedding of the second data when the target construction data includes the first data and the second data;

[0027] The extracting the embedding of the target construction data by using a multimodal embedding model to obtain the construction embedding includes:

[0028] In the case where the target construction data includes the first data, extracting the embedding of the first data using the first embedding model to obtain a first embedding; wherein, in the case where the target construction data only includes the first data, the first embedding is the construction embedding;

[0029] In the case where the target construction data includes the second data, extracting the embedding of the second data using the second embedding model to obtain a second embedding; wherein, in the case where the target construction data only includes the second data, the second embedding is the construction embedding;

[0030] When the target construction data includes first data and second data, the first embedding and the second embedding are fused using a fusion model to obtain the construction embedding.

[0031] In a second aspect, the present application provides an operation risk assessment device, which is used to assess the risks in the process of constructing distribution network equipment. The operation risk assessment device includes:

[0032] An acquisition unit, used to acquire target construction data generated during the construction of the distribution network equipment; the target construction data is data including more than one mode;

[0033] An extraction unit, configured to extract the embedding of the target construction data using a multimodal embedding model to obtain a construction embedding; the multimodal embedding model is configured to extract embeddings of data of more than one modality;

[0034] A risk assessment unit is used to embed the construction into a risk assessment model and determine the target result of the risk assessment through the risk assessment model; the risk assessment model is trained using historical data, and the historical data includes the embedding of historical construction data generated during the construction of the distribution network equipment and historical assessment results, and the historical assessment results are the results of risk assessment based on the historical construction data.

[0035] In a third aspect, the present application provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, computer instructions are stored in the memory, and the processor executes the operational risk management method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0036] In a fourth aspect, the present application provides a computer-readable storage medium, on which a single computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the operational risk management method of the above-mentioned first aspect or any corresponding embodiment thereof.

[0037] In a fifth aspect, the present application provides a computer program product, including computer instructions, which are used to enable a computer to execute the operational risk management method of the above-mentioned first aspect or any corresponding embodiment thereof.

[0038] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application.

[0039] The technical solution provided by this application has the following technical effects:

[0040] In an embodiment of the present application, target construction data generated during the construction of distribution network equipment is obtained; the target construction data is data including more than one modality. The embedding of the target construction data is extracted using a multimodal embedding model as a construction embedding, wherein the multimodal embedding model is used to extract the embedding of data of more than one modality. After obtaining the construction embedding, the operation risk management and control device inputs the construction embedding into the risk assessment model, wherein the risk assessment model is a model trained based on historical data, and the historical data includes the embedding of historical construction data generated during the construction of distribution network equipment and the historical assessment results, and the historical assessment results are the results obtained by risk assessment based on historical construction data. Since the risk assessment model is trained based on the embedding of historical construction data and historical assessment results, the risk assessment model has the ability to perform risk assessment based on construction embedding. After receiving the construction embedding, the risk assessment model performs risk assessment based on the construction embedding to obtain the target result of the risk assessment. Based on the technical solution of the present application, an embedded risk assessment of the target construction data generated during the construction of the distribution network equipment is realized, which solves the problem of being unable to identify data of some modes due to the large number of data modes in the target construction data, avoids the possibility of being unable to conduct risk assessment due to being unable to identify data of some modes, realizes the automation of risk assessment, and improves the efficiency of risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0042] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0043] Figure 1 A schematic diagram of a process flow of an operation risk management method according to an embodiment of the present application;

[0044] Figure 2 This is a schematic diagram of the structure of an operation risk management device according to an embodiment of the present application;

[0045] Figure 3 A schematic diagram of the hardware structure of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are 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 those skilled in the art without making creative work are within the scope of protection of this application.

[0047] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or equipment.

[0048] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0049] The execution subject of the embodiment of the present application is an operation risk control device, wherein the operation risk control device can be any computer device that can execute the technical solution disclosed in the embodiment of the method of the present application. Optionally, the operation risk control device can be one of the following: a computer, a server.

[0050] In an embodiment of the present application, the operation risk management device is used to evaluate the risks in the process of constructing distribution network equipment.

[0051] It should be understood that the method embodiments of the present application may also be implemented by a processor executing computer program code.

[0052] The following describes the embodiments of the present application in conjunction with the accompanying drawings in the embodiments of the present application. Figure 1 , Figure 1 A flowchart of a method for operating risk management and control provided in an embodiment of the present application is shown below. Figure 1 As shown, an operation risk management method is provided in an embodiment of the present application. The operation risk management method is used to evaluate the risks in the process of constructing distribution network equipment. The operation risk management method includes:

[0053] S101: Acquire target construction data generated during the construction of distribution network equipment.

[0054] In an embodiment of the present application, the target construction data is data generated during the construction of the distribution network equipment, and the target construction data is data including more than one modality, wherein the modality represents the data type of the data, and the data of more than one modality represents data of more than one data type. For example, the target construction data includes construction image data during the construction of the distribution network equipment and voltage data generated by the distribution network equipment, wherein the construction image data is picture-type data, and the voltage data is text-type data. For another example, the target construction data includes construction video data and environmental audio data during the construction of the distribution network equipment, wherein the construction video data is picture-type data, and the environmental audio data is audio-type data.

[0055] In an implementation method of acquiring target construction data, the operation risk management and control device receives the target construction data input by an input component. The input component includes at least one of the following: a keyboard, a mouse, a touch screen, a touch pad, and an audio input device.

[0056] In another implementation method of acquiring target construction data, the operation risk management and control device receives the target construction data sent by a terminal. The terminal includes at least one of the following: a mobile phone, a computer, a tablet computer, and a server.

[0057] S102: Extract embedding of target construction data using a multimodal embedding model to obtain construction embedding.

[0058] In an embodiment of the present application, a multimodal embedding model is used to extract embeddings of data of more than one modality, wherein the embedding represents the features of the data, and thus the embedding of the extracted data is the feature of the extracted data. Since the target construction data is data including more than one modality, the multimodal embedding model can be used to extract the embedding of the target construction data to obtain the construction embedding. By extracting the embedding of the target construction data, the construction embedding includes the features of the data of more than one modality in the target construction data.

[0059] In a possible implementation, the multimodal embedding model includes a first embedding model, a second embedding model and a fusion model, the first embedding model is used to extract the embedding of the first data, the first data is data of the first data type, the second embedding model is used to extract the embedding of the second data, the second data is data of the second data type, and the fusion model is used to fuse the embedding of the first data and the embedding of the second data when the target construction data includes the first data and the second data. In the case where the target construction data includes the first data, the embedding of the first data is extracted using the first embedding model to obtain the first embedding. In the case where the target construction data only includes the first data, the first embedding is the construction embedding. In the case where the target construction data includes the second data, the embedding of the second data is extracted using the second embedding model to obtain the second embedding. In the case where the target construction data only includes the second data, the second embedding is the construction embedding. In the case where the target construction data includes the first data and the second data, the first embedding and the second embedding are fused using the fusion model to obtain the construction embedding.

[0060] For example, when the first data type is a picture type, the second data type is a text type, and the target construction data includes the first data of the picture type and the second data of the text type, the first embedding model is used to extract the first data of the picture type to obtain the first embedding, and the second embedding model is used to extract the second data of the text type to obtain the second embedding. After obtaining the first embedding and the second embedding, the first embedding and the second embedding are fused using a fusion model to obtain a construction embedding including the features of the first data and the second data. The embedding of the data of the first data type in the target construction data is extracted using the first embedding model, and the embedding of the data of the second data type in the target construction data is extracted using the second embedding model, and then the embedding output by the first embedding model and the second embedding model are fused using the fusion model. Compared with extracting the embedding of data of multiple data types by one extraction method, the technical solution of the present application sets a multimodal embedding model. By extracting the embedding of different types of data by different embedding models, the accuracy of the extracted construction embedding can be improved, thereby improving the accuracy of risk assessment based on the construction embedding in subsequent steps. It should be understood that the number of models in the multimodal embedding model is related to the number of data types in the target construction data. As an example, the number of models in the multimodal embedding model is equal to the number of data types in the target construction data plus 1. In a case where data of a third data type exists in the target construction data, the multimodal embedding model may include a third embedding model for extracting embedding of the data of the third data type.

[0061] In another possible implementation, a target prompt word can be obtained, and a construction embedding is stored in a multimodal embedding model according to the target prompt word, and then the position of extracting the construction embedding from the multimodal embedding model is determined according to the target prompt word. The target prompt word is a text used to guide the multimodal embedding model to compress the first data and the second data in the target construction data, wherein the target position in the target prompt word is used to write the compressed result, the first data is the data of the first data type in the target construction data, and the second data is the data of the second data type in the target construction data, and the first data type and the second data type are different. After obtaining the target construction data and the target prompt word, the operation risk control device generates a first prompt embedding of the target prompt word in the semantic space based on the first embedding of the first data in the semantic space of the multimodal embedding model and the second embedding of the second data in the semantic space, so that the target position in the first prompt embedding is used to write the embedding of the result obtained by compressing the first data and the second data, wherein the semantic space of the multimodal embedding model is an embedding space, and the embeddings in the semantic space are all embeddings that can be processed by the multimodal embedding model, that is, the information carried by the embedding can be used by the multimodal embedding model. It should be understood that the first prompt embedding is an embedding sequence, and the target prompt word is a word sequence, and the two sequences correspond to each other. The first prompt embedding is encoded using a multimodal embedding model to obtain a target hidden state sequence, so that the hidden state at the target position in the target hidden state sequence is a hidden state obtained by compressing the first data and the second data, that is, the hidden state at the target position in the target hidden state sequence includes information about the first data and information about the second data. Therefore, the operation risk control device can obtain the construction embedding of the target construction data based on the hidden state at the target position in the target hidden state sequence.

[0062] For example, the first data is image b, and the second data is text c. The target prompt word is: "Please compress image b and text c into one word:", where the position after ":" is the target position, that is, the target position is the last position in the target prompt word. The target position is used to write the word obtained by compressing image b and text c, that is, when the compressed word should be written into the target prompt word, the compressed word is the last word in the target prompt word. In the case of generating embedding B of image b and embedding C of text c, a first prompt embedding is generated based on embedding B, embedding C and the target prompt word, and the first prompt embedding includes embedding B and embedding C. After the first prompt embedding is generated, the first prompt embedding is encoded using a multimodal embedding model to obtain a target hidden state sequence. The hidden state at the target position in the target hidden state sequence includes information of embedding B and embedding C, so the hidden state at the target position in the target hidden state sequence can be used as a construction embedding.

[0063] S103: The construction is embedded into the risk assessment model, and the target result of the risk assessment is determined through the risk assessment model.

[0064] In the embodiment of the present application, the risk assessment model is trained using historical data, and the historical data includes the embedding of historical construction data generated during the construction of the distribution network equipment and the historical assessment results, and the historical assessment results are the results of risk assessment based on the historical construction data. Therefore, the risk assessment model has the ability to perform risk assessment based on the construction embedding of the target construction data.

[0065] By inputting the construction embedding into the risk assessment model, the risk assessment model is used to determine the target results of the risk assessment based on the construction embedding. The target results may include the following:

[0066] Risk level: low risk, medium risk, high risk, etc.

[0067] Risk factors: Identify and list specific factors that may cause risks, such as aging equipment, improper operation, harsh environment, etc.

[0068] Risk area: Specifies the specific area or construction scope affected by the risk.

[0069] Risk causes: Analyze the specific causes of the risks, such as equipment abnormalities, environmental changes, etc.

[0070] Risk description: A detailed description of the nature of the risk, its possible impact and consequences.

[0071] Risk probability: Assess the likelihood of a risk occurring, which can be a qualitative description or a quantitative probability value.

[0072] Risk impact: Assess the potential impact of risks on personnel safety, equipment operation, production efficiency, etc.

[0073] Risk source location: Determine the specific location or link of the risk source to facilitate the adoption of targeted control measures.

[0074] Risk Control Recommendations: Suggested measures to reduce or control risks.

[0075] In order to conduct risk assessment on the construction of distribution network equipment, the usual practice is to use the target construction data generated during the construction of distribution network equipment as the input of the traditional model, and use the traditional model to conduct risk assessment based on the target construction data. However, since the modalities of the data generated by different distribution network equipment may be different, if the risk assessment work of the traditional model needs to cover all distribution network equipment, the traditional model needs data of all modalities generated by the construction of all distribution network equipment during the training process. For example, if the data of modal A generated by the construction of distribution network equipment is missing during the training process of the traditional model, the traditional model cannot conduct risk assessment on the data of modal A generated during the construction of distribution network equipment. This imposes great limitations on the risk assessment of the traditional model.

[0076] In the embodiment of the present application, the risk assessment model is obtained by embedding training based on the historical construction data generated during the construction of the distribution network equipment. Therefore, the risk assessment model can be based on the embedding of any modality of data generated by the construction of any distribution network equipment to perform risk assessment, solving the problem that the traditional model has large limitations in risk assessment due to different modalities of data.

[0077] In an embodiment of the present application, the operation risk control device obtains the target construction data generated in the process of constructing the distribution network equipment, wherein the target construction data is data including more than one modality. The embedding of the target construction data is extracted using a multimodal embedding model as a construction embedding, wherein the multimodal embedding model is used to extract the embedding of more than one modality of data. After obtaining the construction embedding, the operation risk control device inputs the construction embedding into the risk assessment model, wherein the risk assessment model is a model trained based on historical data, and the historical data includes the embedding of historical construction data generated in the process of previous construction of the distribution network equipment and the historical assessment results, and the historical assessment results are the results of risk assessment based on historical construction data. Since the risk assessment model is trained based on the embedding of historical construction data and historical assessment results, the risk assessment model has the ability to perform risk assessment based on construction embedding. After receiving the construction embedding, the risk assessment model performs risk assessment based on the construction embedding to obtain the target result of the risk assessment. Through the above steps, the operation risk management and control device realizes the embedded risk assessment of the target construction data generated during the construction process of the distribution network equipment, reduces the possibility of being unable to identify some modal data due to the large number of data modalities in the target construction data, and thus being unable to perform risk assessment, realizes the automation of risk assessment, and improves the efficiency of risk assessment.

[0078] In combination with any embodiment of the present application, the obtaining of target construction data generated during the construction of the distribution network equipment in S101 specifically includes:

[0079] S1011: Using a collection device to collect target construction data generated during the construction of distribution network equipment.

[0080] In an embodiment of the present application, the collection device is a device for collecting data generated during the construction of distribution network equipment. For example, the collection device includes a handheld terminal and a communication terminal. The target construction data includes personnel information data, construction equipment data, and construction environment data. The personnel information data includes: information about construction personnel who construct the distribution network equipment, and the construction equipment data includes: data generated by the distribution network equipment during the construction of the distribution network equipment. The construction environment data represents data about the environment during the construction of the distribution network equipment. The handheld terminal is used to collect and store personnel information data, and the communication terminal is used to collect and store construction equipment data and construction environment data.

[0081] When the collection device is a handheld terminal, the personnel information data of the construction personnel is collected by the handheld terminal. Optionally, the handheld terminal is developed based on application requirements. For example, a handheld terminal integrating scanning code, taking photos, fingerprint recognition, electronic signature, and positioning is developed. The handheld terminal supports the functions of face recognition, fingerprint collection and verification, as well as the functions of construction personnel at all levels performing their duties, construction personnel clocking in to start work, and construction personnel qualification verification. The above functions can be realized through mobile applications.

[0082] When the acquisition device is a communication terminal, the communication terminal is used to collect the construction environment data and the data generated by the distribution network equipment during the construction process of the distribution network equipment. Optionally, the communication terminal is developed based on the application requirements. For example, the communication terminal such as the law enforcement instrument, the control ball, the safety helmet, etc. that integrates law enforcement records and intercom calls is developed. The communication terminal can collect real-time data. For example, the video data collection and capture of the construction site. Another example is the monitoring and collection of the data generated by the distribution network equipment during the construction process of the distribution network equipment.

[0083] In the case where the target construction data includes personnel information data, the construction embedding includes the features of the personnel information data. In the case where the construction embedding includes the features of the personnel information data, the risk assessment model can be used to determine the construction risks of the construction personnel based on the features of the personnel information data. For example, in the case where the personnel information data includes a construction personnel picture of the construction personnel, the risk assessment model can be used to determine the construction risks of the construction personnel not wearing a safety helmet, losing the protection of the safety belt, etc. based on the features of the construction personnel picture. For another example, in the case where the personnel information includes the construction qualifications of the construction personnel, the risk assessment model can be used to determine the construction risks of the construction personnel based on the construction qualifications of the construction personnel, such as the construction qualifications of the construction personnel not meeting the construction conditions.

[0084] In the case where the target construction data includes construction equipment data, the construction embedding includes the features of the construction equipment data. In the case where the construction embedding includes the features of the construction equipment data, the risk assessment model can be used to determine the construction risks generated by the distribution network equipment based on the features of the construction equipment data. For example, in the case where the construction equipment data includes the voltage data of the distribution network equipment, the risk assessment model can be used to determine the construction risks such as voltage anomalies in the distribution network equipment based on the voltage data of the distribution network equipment. For another example, in the case where the construction equipment data includes the equipment load data of the distribution network equipment, the risk assessment model can be used to determine the construction risks such as load imbalance and overload in the distribution network equipment based on the equipment load data of the distribution network equipment.

[0085] In the case where the target construction data includes construction environment data, the construction embedding includes the features of the construction environment data. In the case where the construction embedding includes the features of the construction environment data, the risk assessment model can be used to determine the construction risks that exist in the environment where the distribution network is constructed based on the features of the construction environment data. For example, in the case where the construction environment data includes ambient temperature data, the risk assessment model can be used to determine, based on the ambient temperature data, that there are construction risks in the environment where the distribution network equipment is constructed, such as excessively high temperatures that may cause equipment overheating. For another example, in the case where the construction environment data includes ambient humidity data, the risk assessment model can be used to determine, based on the ambient humidity data, that there are construction risks in the environment where the distribution network equipment is constructed, such as excessively high humidity that may cause short circuits in the distribution network equipment, leakages in the distribution network equipment, etc.

[0086] Optionally, the target construction data includes associated data generated by associated equipment during the construction of the distribution network equipment, wherein the associated equipment is equipment other than the distribution network equipment in the circuit where the distribution network is located. By collecting the associated data generated by the associated equipment during the construction process, the risk assessment model can be used to assess the risks caused to the associated equipment during the construction of the distribution network equipment, thereby achieving operational risk management and control of the associated equipment in the distribution network equipment circuit.

[0087] In this embodiment, the operation risk control device can use the collection equipment to collect the target construction data generated during the construction of the distribution network equipment, wherein the collection equipment includes but is not limited to a handheld terminal and a communication terminal. Different target construction data can be collected through different collection equipment to obtain richer target construction data, for example, collecting personnel information data through a handheld terminal. For another example, construction equipment data and construction environment data are collected through a communication terminal. The richer the information of the target construction data, the higher the accuracy of the target result of risk assessment based on the embedding of the target construction data using the risk assessment model.

[0088] In combination with any embodiment of the present application, the step of extracting the embedding of the target construction data using the multimodal embedding model in S102 to obtain the construction embedding specifically includes:

[0089] S1021: Extract the embedding of the target construction data using the multimodal embedding model to obtain a candidate embedding. In an embodiment of the present application, the operation risk management and control device extracts the embedding of the target construction data using the multimodal embedding model, that is, extracts features of more than one modality of data in the target construction data to obtain a candidate embedding.

[0090] S1022: Split the candidate embeddings into n groups of reference embeddings.

[0091] In an embodiment of the present application, the n groups of reference embeddings are n groups of embeddings obtained by dividing the candidate embeddings, and the n groups of reference embeddings include a first embedding, and the first embedding includes m first vectors, where n is a positive integer greater than 1. For example, when the candidate embedding is [x1, x2, x3], the candidate embedding is divided into a first embedding [x1, x2] and an embedding [x3]. x1 and x2 in the first embedding represent the m first vectors in the first embedding.

[0092] S1023: Determine j second vectors for representing the m first vectors from the first embedding.

[0093] In the embodiment of the present application, j is greater than or equal to 1, m is a positive integer greater than j and greater than 1, and the second vector is used to represent the m first vectors. The second vector representing the m first vectors is obtained by selecting or generating j vectors from the m first vectors.

[0094] In a possible implementation, j second vectors are determined according to the first embedding (m first vectors), the m first vectors may be divided into j clusters, and the central vectors of the j clusters are used as the j second vectors.

[0095] In another possible implementation, j second vectors are determined based on the first embedding (m first vectors). The m first vectors can be divided into j groups of candidate vectors, and the average value of the vectors in any group of candidate vectors is calculated to obtain j candidate means corresponding to the j groups, and the j candidate means are used as j second vectors.

[0096] S1024: Determine the construction embedding according to the j second vectors and the third vector.

[0097] In an embodiment of the present application, the third vector is used to characterize the second embedding, and the second embedding is an embedding other than the first embedding in the n sets of reference embeddings. The construction embedding includes a third vector and j second vectors, wherein the third vector is used to characterize the second embedding, and the second embedding is an embedding other than the first embedding in the n sets of reference embeddings. For example, the candidate embedding is divided into the first embedding [x1, x2], the embedding [x3], and the embedding [x4], wherein the embedding [x3] and the embedding [x4] are the second embeddings.

[0098] In the embodiment of the present application, by reducing the m first vectors in the first embedding to j second vectors, the number of vectors in the target construction embedding can be reduced. When the number of vectors in the construction embedding is reduced, the time for the risk assessment model to process the parameters in the construction embedding can be reduced, thereby increasing the speed of determining the target result based on the construction embedding using the risk assessment model.

[0099] Optionally, the third vector is obtained by reducing the number of vectors in the second embedding. Specifically, k fourth vectors in the second embedding are determined, and L third vectors used to characterize the k fourth vectors are determined from the second embedding. In one possible implementation, the k fourth vectors are divided into L clusters, and the center vectors of the L clusters are used as the L third vectors. In another possible implementation, the k fourth vectors are divided into L groups of reference vectors, and the average value of the vectors in any group of reference vectors is calculated to obtain L reference means corresponding to the L groups of reference vectors, and the L reference means are used as the L third vectors.

[0100] In this application embodiment, the operation risk management and control device extracts the embedding of the target construction data by using a multimodal embedding model to obtain a candidate embedding, and divides the candidate embedding into n groups of reference embeddings, wherein the n groups of reference embeddings include a first embedding, and the first embedding includes m first vectors, wherein m and n are both positive integers greater than 1. After determining the m first vectors in the first embedding, j second vectors for characterizing the m first vectors are determined, thereby reducing the amount of parameters in the first embedding. When the j second vectors are determined, the construction embedding is determined based on the j second vectors and the third vector, wherein the third vector is used to characterize the second embedding other than the first embedding in the n groups of reference embeddings. The above steps can reduce the amount of parameters in the construction embedding, thereby reducing the time for the risk assessment model to process the parameters in the construction embedding, and improving the speed of determining the target result based on the construction embedding using the risk assessment model.

[0101] In conjunction with any embodiment of the present application, the target result includes a target risk level of the construction risk of the current construction determined by the risk assessment model. After the construction is embedded into the risk assessment model and the target result of the risk assessment is determined by the risk assessment model, the operation risk management method further includes:

[0102] S201: When the target risk level does not meet the preset warning level, the construction embedding is input into the trend prediction model, and the embedding change trend of the construction embedding is predicted by the trend prediction model.

[0103] In the embodiment of the present application, the embedding change trend is the change trend of the construction embedding after the current moment.

[0104] In an embodiment of the present application, the target risk level represents the possibility of risks occurring during the current construction process, and the preset warning level represents the level at which construction personnel need to take preventive measures to reduce the possibility of construction risks occurring. When the target risk level does not meet the preset warning level, it means that at the current moment, the possibility of risks occurring in this construction is low, so construction personnel may not take preventive measures to reduce the probability of construction risks occurring. For example, the risk levels include level one and level two. When the preset warning level is level two and the target risk level is level one, the target risk level does not meet the preset warning level, and at the current moment, the possibility of risks occurring in this construction is low.

[0105] In an embodiment of the present application, when the target risk level does not meet the preset warning level, the construction embedding is input into the trend prediction model, and the embedding change trend of the construction embedding is predicted by the trend prediction model, wherein the embedding change trend is the change trend of the construction embedding after the current moment. In a possible implementation, the construction embedding is the embedding of the target construction data acquired at the first moment (the current moment), and the embedding change trend is the change trend of the construction embedding from the first moment to the second moment. The time interval between the first moment and the second moment is a preset time length, and the preset time length can be set and modified according to actual needs. For example, the target construction data includes the equipment load data of the distribution network equipment at the first moment, and the construction embedding includes the characteristics of the equipment load data, and the equipment load data is 50%. The trend prediction model is used to predict that the change trend of the equipment load data in the embedding change trend from the first moment to the second moment is an increase of 40%, indicating that the trend prediction model predicts that the equipment load data of the distribution network equipment at the second moment is 90%.

[0106] S202: Determine the prediction result of the construction risk generated by the current construction after the current moment according to the embedded change trend.

[0107] In the embodiment of the present application, the embedding change trend represents the change trend of the construction embedding from the first moment to the second moment. Using the embedding change trend, the prediction result of the construction risk generated after the current moment can be predicted. The prediction result may include the following contents:

[0108] Risk level: low risk, medium risk, high risk, etc.

[0109] Risk factors: Identify and list specific factors that may lead to construction risks, such as aging equipment, improper operation, harsh environment, etc.

[0110] Risk causes: Analyze the specific causes that lead to construction risks, such as equipment abnormalities, environmental changes, etc.

[0111] Risk description: Provide a detailed description of the nature, possible impacts and consequences of the construction risk.

[0112] Risk probability: Assess the possibility of construction risk occurring, which can be a qualitative description or a quantitative probability value.

[0113] Risk impact: Assess the potential impact of construction risks on personnel safety, equipment operation, production efficiency, etc.

[0114] Risk Control Recommendations: Recommended measures to reduce or control construction risks.

[0115] In a possible implementation, a prediction result corresponding to the embedded change trend is determined from the preset matching result. For example, there is a matching relationship between abnormal voltage fluctuation and excessive equipment load in the preset matching result. In the embedded change trend, the voltage of the distribution network rises by more than 100%, indicating that the voltage has abnormal fluctuations. In the case that the embedded change trend has abnormal voltage fluctuations, the prediction result includes excessive equipment load.

[0116] In another possible implementation, the embedding change trend is input into the risk assessment model, and the risk assessment model is used to perform risk assessment on the distribution network equipment after the current moment to obtain a prediction result. For example, the construction embedding includes the characteristics of the construction equipment data generated by the distribution network equipment at the first moment during the construction of the distribution network equipment, and the embedding change trend is the change trend of the construction embedding from the first moment to the second moment, that is, the change trend of the construction equipment data from the first moment to the second moment. The risk assessment model is used to perform risk assessment based on the change trend of the construction equipment data from the first moment to the second moment, and the construction risk existing from the first moment to the second moment during the construction of the distribution network equipment is predicted to obtain a prediction result.

[0117] S203: Send the prediction results to construction personnel who are constructing the distribution network equipment, so that the construction personnel can check the construction risks according to the prediction results during the construction process.

[0118] In an embodiment of the present application, after the prediction result is determined, the prediction result is sent to the construction personnel who are constructing the distribution network equipment, and the construction personnel can conduct investigations based on the prediction result.

[0119] In one possible implementation, when the prediction results include abnormal voltage fluctuations in distribution network equipment after the current moment, construction personnel can reduce the equipment connected to the distribution network equipment, thereby reducing the possibility of abnormal voltage fluctuations in the distribution network equipment by reducing the load pressure of the distribution network equipment.

[0120] In another possible implementation, when the prediction results include that the ambient humidity is too high after the current moment, causing a short circuit in the distribution network equipment, the construction personnel can use dehumidification equipment to reduce the humidity at the construction site, thereby reducing the possibility of a short circuit in the distribution network equipment due to excessive ambient humidity.

[0121] Optionally, the measurement results include risk control suggestions for the predicted construction risks, that is, recommended measures to reduce or control risks. Construction personnel can check the construction risks according to the risk control suggestions to reduce the probability of construction risks occurring after the current moment.

[0122] In this embodiment, when the target risk level does not meet the preset warning level, that is, when the possibility of construction risk at the current moment is low, the operation risk control device uses the trend prediction model to predict the embedded change trend of the construction embedding to obtain the change trend of the construction embedding after the current moment, and uses the change trend to determine the prediction result of the construction risk generated by this construction after the current moment, and sends the prediction result to the construction personnel who are constructing the distribution network equipment, so that the construction personnel can check the distribution network equipment according to the prediction result. By checking the distribution network equipment, potential construction risks can be discovered in time, and the probability of construction risks after the current moment can be reduced.

[0123] In conjunction with any embodiment of the present application, the target result includes a target risk level of the construction risk of the current construction determined by the risk assessment model. After the construction is embedded into the risk assessment model and the target result of the risk assessment is determined by the risk assessment model, the operation risk management method further includes:

[0124] S301: When the target risk level meets the preset warning level, the target result is sent to the construction personnel who are constructing the distribution network equipment, so that the construction personnel can rectify the construction risks according to the target result.

[0125] In the embodiment of the present application, when the target risk level meets the preset warning level, it means that at the current moment, the possibility of construction risk incurred in this construction is high, so the construction personnel should take preventive measures to reduce the probability of construction risk. For example, the risk level includes level one and level two. When the preset warning level is level two and the target risk level is level two, the target risk level meets the preset warning level, and at the current moment, the possibility of risk incurred in this construction is high, and the construction personnel should take preventive measures to reduce the possibility of construction risk.

[0126] In an embodiment of the present application, when the target risk level meets the preset warning level, that is, at the current moment, the possibility of construction risk incurred by this construction is high, the target result is sent to the construction personnel who are constructing the distribution network equipment, so that the construction personnel can make rectifications according to the target result.

[0127] In one possible implementation, when the target result includes that the distribution network equipment has tripped too many times at the current moment and the target risk level meets the preset warning level, the construction personnel can investigate the reasons for the adjustment of the distribution network equipment, thereby reducing the possibility of the distribution network equipment tripping.

[0128] In another possible implementation, when the target result includes the occurrence of thunderstorms in the construction environment and the target risk level meets the preset warning level, construction personnel can cut off the power supply to the distribution network equipment to reduce the possibility of damage to the distribution network equipment by the current generated by lightning.

[0129] S302: Obtaining rectification data generated by construction personnel in the process of rectifying construction risks according to target results.

[0130] In the embodiment of the present application, after the construction personnel complete the rectification according to the target result, the rectification data generated in the process of rectifying the distribution network equipment is obtained. It should be understood that the rectification data obtained in this step has the same data type as the target construction data obtained in S101, so the rectification data also includes more than one modality.

[0131] S303: Extract the embedding of the rectification data using a multimodal embedding model to obtain a rectification embedding.

[0132] In the embodiment of the present application, the multimodal embedding model is used to extract the embedding of data of more than one modality, wherein the embedding represents the features of the data, and thus the embedding of the extracted data is the feature of the extracted data. Since the rectified data is data including more than one modality, the embedding of the rectified data can be extracted using the multimodal embedding model to obtain the rectified embedding. By extracting the embedding of the rectified data, the rectified embedding includes the features of the data of more than one modality in the rectified data.

[0133] S304: embed the rectification into the risk assessment model, and determine the rectification result of the risk assessment through the risk assessment model.

[0134] In the embodiment of the present application, by embedding the rectification into the risk assessment model, the risk assessment model can be used to determine the rectification results of the risk assessment after the construction personnel complete the rectification based on the rectification embedding.

[0135] In the embodiment of the present application, when the target risk level of the construction risk meets the preset warning level, that is, the construction risk is likely to be generated in this construction, the target result is sent to the construction personnel, and the construction personnel are required to make rectifications for the construction risk. After the rectification is completed, the rectification data generated during the construction of the distribution network equipment is obtained, and the risk assessment model is used to perform risk assessment based on the rectification embedding of the rectification data to obtain the rectification result.

[0136] In a possible implementation, the risk level of the above-mentioned construction risk in the rectification result can be determined through the rectification result. If the risk level of the above-mentioned construction risk in the rectification result meets the preset warning level, it means that the rectification of the construction personnel has not solved the construction risk in the target result, and the construction personnel should be urged to make rectification again based on the rectification result. If the risk level of the above-mentioned construction risk in the rectification result does not meet the preset warning level, it means that the rectification of the construction personnel has solved the construction risk in the target result.

[0137] In another possible implementation scenario, the target result is the result of risk assessment on the data generated at the first moment during the construction of the distribution network equipment, and the rectification result is the result of risk assessment on the data generated at the second moment during the construction of the distribution network equipment. The target result and the rectification result are input into the trend prediction model to obtain the rectification change trend of the construction risk from the second moment to the third moment. When the rectification change trend of the construction risk tends to decrease, it means that the rectification of the construction personnel has solved the construction risk in the target result.

[0138] In another possible implementation scenario, the rectification effect of the construction personnel can be determined by comparing the rectification result with the target result. Specifically, the rectification similarity between the rectification result and the target result is calculated. If the rectification similarity is greater than the rectification threshold, it means that the rectification of the construction personnel has resolved the construction risk in the target result.

[0139] In this embodiment, when the target risk level of the construction risk meets the preset warning level, that is, when the possibility of construction risk at the current moment is high, the operation risk control device sends the target result to the construction personnel, so that the construction personnel can rectify the construction risk according to the target result, and after the rectification is completed, obtain the rectification data generated in the process of construction of the distribution network equipment, and use the risk assessment model to perform risk assessment based on the rectification embedding of the rectification data to obtain the rectification result. Based on the rectification result determined based on the rectification data after rectification, the rectification effect of the rectification carried out by the construction personnel according to the target result can be determined. Optionally, in the case where the rectification effect is not ideal, the operation risk control device can reduce the risk level of the construction risk by repeating S301 to S304 until the rectification effect is ideal, thereby ensuring the effectiveness of the management of the construction risk.

[0140] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0141] It should be noted that the contents not described in detail in the present application specification belong to the common knowledge of those skilled in the art.

[0142] In this embodiment, a work risk management device is also provided, and a single device is used to implement the above-mentioned embodiments and optional implementation methods, which have been described and will not be repeated. As used below, the term "unit" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0143] Figure 2 It is a schematic diagram of the structure of an operation risk management device according to an embodiment of the present application.

[0144] This application provides an operation risk management device, such as Figure 2 As shown, the operation risk control device 10 is used to evaluate the risks in the process of constructing the distribution network equipment. The operation risk control device 10 includes: an acquisition unit 11, an extraction unit 12, and a risk assessment unit 13. Among them:

[0145] The acquisition unit 11 is used to acquire target construction data generated during the construction of the distribution network equipment. The target construction data is data including more than one mode, and the mode represents the data type of the data.

[0146] The extraction unit 12 is used to extract the embedding of the target construction data using the multimodal embedding model to obtain the construction embedding. The multimodal embedding model is used to extract the embedding of data of more than one modality.

[0147] The risk assessment unit 13 is used to embed the construction into the risk assessment model, and determine the target result of the risk assessment through the risk assessment model. The risk assessment model is trained using historical data, and the historical data includes the embedding of historical construction data generated during the construction of the distribution network equipment and historical assessment results, and the historical assessment results are the results of risk assessment based on the historical construction data.

[0148] In combination with any embodiment of the present application, the acquisition unit 11 is specifically used for:

[0149] The target construction data generated during the construction of the distribution network equipment is collected by using the collection equipment. The target construction data includes personnel information data, construction equipment data, and construction environment data. The personnel information data includes: information of the construction personnel who construct the distribution network equipment. The construction equipment data includes: data generated by the distribution network equipment during the construction of the distribution network equipment. The construction environment data represents the data of the environment during the construction of the distribution network equipment. The collection equipment includes a handheld terminal and a communication terminal. The handheld terminal is used to collect and store personnel information data, and the communication terminal is used to collect and store construction equipment data and construction environment data.

[0150] In combination with any embodiment of the present application, the extraction unit 12 is specifically used for:

[0151] The embedding of the target construction data is extracted using a multimodal embedding model to obtain a candidate embedding. The candidate embedding is split to obtain n groups of reference embeddings. The n groups of reference embeddings include a first embedding, and the first embedding includes m first vectors, where n is a positive integer greater than 1. From the first embedding, j second vectors are determined to represent the m first vectors. j is greater than or equal to 1, and m is a positive integer greater than j and greater than 1. The construction embedding is determined based on the j second vectors and the third vector. The third vector is used to represent the second embedding, and the second embedding is the embedding in the n groups of reference embeddings except the first embedding.

[0152] In conjunction with any implementation method of the present application, the target result includes a target risk level of the construction risk of the current construction determined by the risk assessment model, such as Figure 2 As shown, the above-mentioned operation risk control device 10 also includes a prediction unit 14, and the prediction unit 14 is used to:

[0153] When the target risk level does not meet the preset warning level, the construction embedding is input into the trend prediction model, and the embedding change trend of the construction embedding is predicted by the trend prediction model. The embedding change trend is the change trend of the construction embedding after the current moment. The prediction result of the construction risk generated by this construction after the current moment is determined based on the embedding change trend. The prediction result is sent to the construction personnel who are constructing the distribution network equipment, so that the construction personnel can check the construction risks according to the prediction results during the construction process.

[0154] In combination with any implementation manner of the present application, Figure 2 As shown, the above-mentioned operation risk control device 10 also includes a rectification unit 15, and the rectification unit 15 is used to:

[0155] When the target risk level meets the preset warning level, the target result is sent to the construction personnel who are constructing the distribution network equipment, so that the construction personnel can rectify the construction risk according to the target result. The rectification data generated in the process of the construction personnel rectifying the construction risk according to the target result is obtained. The embedding of the rectification data is extracted using the multimodal embedding model to obtain the rectification embedding. The rectification embedding is input into the risk assessment model, and the rectification result of the risk assessment is determined by the risk assessment model.

[0156] In combination with any embodiment of the present application, the modal embedding model includes a first embedding model, a second embedding model and a fusion model, the first embedding model is used to extract the embedding of first data, the first data is data of a first data type, the second embedding model is used to extract the embedding of second data, the second data is data of a second data type, and the fusion model is used to fuse the embedding of the first data and the embedding of the second data when the target construction data includes the first data and the second data.

[0157] The extraction unit 12 is further used for: when the target construction data includes the first data, extracting the embedding of the first data using the first embedding model to obtain the first embedding. Wherein, when the target construction data only includes the first data, the first embedding is the construction embedding. When the target construction data includes the second data, extracting the embedding of the second data using the second embedding model to obtain the second embedding. Wherein, when the target construction data only includes the second data, the second embedding is the construction embedding. When the target construction data includes the first data and the second data, the first embedding and the second embedding are fused using the fusion model to obtain the construction embedding.

[0158] In some embodiments, the functions or units included in the device provided in the embodiments of the present application can be used to execute the method described in the above method embodiment. Its specific implementation can refer to the description of the above method embodiment. For the sake of brevity, it will not be repeated here. The operation risk management device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0159] Figure 3 A schematic diagram of the hardware structure of a computer device provided in an embodiment of the present application. The computer device 20 includes a processor 21 and a memory 22. Optionally, the computer device 20 also includes an input device 23 and an output device 24. The processor 21, the memory 22, the input device 23 and the output device 24 are coupled via a connector, and the connector includes various interfaces, transmission lines or buses, etc., which are not limited in the embodiments of the present application. It should be understood that in each embodiment of the present application, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, it can be connected through various interfaces, transmission lines, buses, etc.

[0160] The processor 21 may include one or more processors, for example, one or more central processing units (CPUs). When the processor is a CPU, the CPU may be a single-core CPU or a multi-core CPU. Optionally, the processor 21 may be a processor group consisting of multiple CPUs, and the multiple processors are coupled to each other via one or more buses. Optionally, the processor may also be other types of processors, etc., which are not limited in the embodiments of the present application.

[0161] The memory 22 can be used to store computer program instructions and various computer program codes including the program code for executing the program code of the present application. Optionally, the memory includes but is not limited to random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read only memory (CD-ROM), which is used for related instructions and data.

[0162] The input device 23 is used to input data and / or signals, and the output device 24 is used to output data and / or signals. The input device 23 and the output device 24 can be independent devices or an integrated device.

[0163] It can be understood that in the embodiment of the present application, the memory 22 can be used not only to store relevant instructions, but also to store relevant data. For example, the memory 22 can be used to store target construction data obtained through the input device 23, or the memory 22 can also be used to store target results obtained through the processor 21, etc. The embodiment of the present application does not limit the specific data stored in the memory.

[0164] Understandably, Figure 3 Only a simplified design of a computer device is shown. In practical applications, the computer device may also include other necessary components, including but not limited to any number of input / output devices, processors, memories, etc., and all computer devices that can implement the embodiments of the present application are within the protection scope of the present application.

[0165] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0166] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. Those skilled in the art can also clearly understand that the descriptions of the various embodiments of the present application have different focuses. For the convenience and brevity of description, the same or similar parts may not be repeated in different embodiments. Therefore, for parts not described or not described in detail in a certain embodiment, refer to the records of other embodiments.

[0167] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0168] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0169] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0170] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted by a computer-readable storage medium. The computer instructions can be transmitted from a website site, a computer, a server or a data center to another website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server, a data center, etc. that contains one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0171] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by a computer program to instruct the relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned method embodiments. The aforementioned storage medium includes: a read-only memory (ROM) or a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

Claims

1. A method for operation risk management and control, characterized in that: The method is used to assess the risk during the construction of distribution network equipment, and the method comprises: Acquire target construction data generated during the construction of the distribution network equipment; the target construction data is data including more than one mode, the data type of the mode characterizing the data; Extracting the embedding of the target construction data using a multimodal embedding model to obtain a construction embedding; the multimodal embedding model is used to extract embedding of data of more than one modality; The construction is embedded into a risk assessment model, and the target result of the risk assessment is determined through the risk assessment model; the risk assessment model is trained using historical data, and the historical data includes the embedding of historical construction data generated during the construction of the distribution network equipment and historical assessment results, and the historical assessment results are the results of risk assessment based on the historical construction data.

2. The method according to claim 1, characterized in that: The obtaining of target construction data generated during the construction of the distribution network equipment includes: The target construction data generated during the construction of the distribution network equipment is collected by using a collection device; the target construction data includes personnel information data, construction equipment data, and construction environment data, the personnel information data includes: information of construction personnel who construct the distribution network equipment, the construction equipment data includes: data generated by the distribution network equipment during the construction of the distribution network equipment; the construction environment data represents data of the environment during the construction of the distribution network equipment, the collection device includes a handheld terminal and a communication terminal, the handheld terminal is used to collect and store the personnel information data, and the communication terminal is used to collect and store the construction equipment data and the construction environment data.

3. The method according to claim 1 or 2, characterized in that: The extracting the embedding of the target construction data by using a multimodal embedding model to obtain the construction embedding includes: extracting embeddings of the target construction data using the multimodal embedding model to obtain candidate embeddings; The candidate embedding is divided into n groups to obtain n reference embeddings; the n groups of reference embeddings include a first embedding, the first embedding includes m first vectors, and n is a positive integer greater than 1; Determine j second vectors for representing m first vectors from the first embedding; j is greater than or equal to 1, and m is a positive integer greater than j and greater than 1; The construction embedding is determined based on j second vectors and third vectors; the third vector is used to represent a second embedding, and the second embedding is an embedding other than the first embedding in the n groups of reference embeddings.

4. The method according to claim 1, characterized in that The target result includes a target risk level of the construction risk of the current construction determined by the risk assessment model, and the method further includes: In the case where the target risk level does not meet the preset warning level, the construction embedding is input into a trend prediction model, and the embedding change trend of the construction embedding is predicted by the trend prediction model; the embedding change trend is the change trend of the construction embedding after the current moment; Determine, according to the embedded change trend, a prediction result of the construction risk generated by the current construction after the current moment; The prediction results are sent to construction personnel who are constructing the distribution network equipment, so that the construction personnel can check the construction risks according to the prediction results during the construction process.

5. The method according to claim 4, characterized in that The method further comprises: When the target risk level meets the preset warning level, the target result is sent to the construction personnel who are constructing the distribution network equipment, so that the construction personnel can rectify the construction risk according to the target result; Obtaining rectification data generated by the construction personnel in the process of rectifying the construction risks according to the target results; extracting an embedding of the rectification data using the multimodal embedding model to obtain a rectification embedding; The rectification is embedded into the risk assessment model, and the rectification result of the risk assessment is determined by the risk assessment model.

6. The method according to claim 1 or 2, characterized in that: The modal embedding model includes a first embedding model, a second embedding model and a fusion model, the first embedding model is used to extract the embedding of first data, the first data is data of a first data type, the second embedding model is used to extract the embedding of second data, the second data is data of a second data type, and the fusion model is used to fuse the embedding of the first data with the embedding of the second data when the target construction data includes the first data and the second data; The extracting the embedding of the target construction data by using a multimodal embedding model to obtain the construction embedding includes: In the case where the target construction data includes the first data, extracting the embedding of the first data using the first embedding model to obtain a first embedding; wherein, in the case where the target construction data only includes the first data, the first embedding is the construction embedding; In the case where the target construction data includes the second data, extracting the embedding of the second data using the second embedding model to obtain a second embedding; wherein, in the case where the target construction data only includes the second data, the second embedding is the construction embedding; When the target construction data includes first data and second data, the first embedding and the second embedding are fused using a fusion model to obtain the construction embedding.

7. An operation risk management device, characterized in that: The device is used to assess the risks in the process of constructing distribution network equipment, and the operation risk control device includes: An acquisition unit, used to acquire target construction data generated during the construction of the distribution network equipment; the target construction data is data including more than one mode, the data type of the mode characterizing the data; An extraction unit, configured to extract the embedding of the target construction data using a multimodal embedding model to obtain a construction embedding; the multimodal embedding model is configured to extract embeddings of data of more than one modality; A risk assessment unit is used to embed the construction into a risk assessment model and determine the target result of the risk assessment through the risk assessment model; the risk assessment model is trained using historical data, and the historical data includes the embedding of historical construction data generated during the construction of the distribution network equipment and historical assessment results, and the historical assessment results are the results of risk assessment based on the historical construction data.

8. A computer device, characterized in that: include: A processor and a memory, wherein the memory is used to store computer program code, wherein the computer program code includes computer instructions, and when the processor executes the computer instructions, the computer device executes the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor is enabled to execute the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The computer program product comprises a computer program or instructions; when the computer program or instructions are run on a computer, the computer is enabled to execute the method according to any one of claims 1 to 6.

Images