Data processing method and system based on power generation enterprise safety risk intelligent prediction model

Abstract

The invention provides a data processing method based on a power generation enterprise safety risk intelligent prediction model, and belongs to the technical field of data processing. First monitoring information is inputted to a trained neural network training model, first risk information and a first risk index are acquired according to an output result of the neural network model, and then the first risk information and the first risk index are acquired; the first risk index is compared with a preset threshold value, if the first risk index is larger than or equal to the preset threshold value, second monitoring information with the information granularity being second granularity is obtained in a second period, the second granularity is smaller than the first granularity, and finally the second monitoring information is input into a trained neural network training model, and obtaining second risk information and a second risk index according to an output result of the neural network model.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a data processing method and system based on an intelligent prediction model for safety risks of power generation enterprises. Background Art

[0002] As power generation companies expand in size and their operations become more complex, the management and control of safety risks become increasingly important. Traditional safety risk assessment methods rely on manual experience and regular inspections, which is not only inefficient but also difficult to fully and accurately capture potential safety risks. Therefore, how to use advanced data processing technology and intelligent algorithms to achieve intelligent prediction and real-time monitoring of safety risks in power generation companies has become an urgent problem to be solved.

[0003] At present, although some safety risk prediction models based on data analysis and machine learning algorithms have been proposed, these models still face many problems in practical applications. For example, the model has high requirements for data quality. When the data granularity is large, the corresponding data granularity of the predicted result data is also large. When importing the model, high-precision data can be used to obtain more detailed prediction results, but the computing power of the processing equipment is required for a long time. At the same time, obtaining high-precision detection data for a long time also increases labor and equipment costs. Summary of the invention

[0004] The embodiments of the present application provide a data processing method and system based on an intelligent prediction model of safety risks of power generation enterprises to improve the above-mentioned problems.

[0005] In order to achieve the above objectives, this application adopts the following technical solutions: In the first aspect, the embodiment of the present application proposes a data processing method based on an intelligent prediction model of safety risks of power generation enterprises, which is applicable to a data processing system based on an intelligent prediction model of safety risks of power generation enterprises. The data processing system based on the intelligent prediction model of safety risks of power generation enterprises includes a controller, and the method is applicable to the controller, including: Obtaining a training set, and using the training set to train a neural network training model, the neural network training model is used to input monitoring information and input risk index, and output a predicted risk index and predicted risk information; Acquire first monitoring information with a first information granularity in a first monitoring period, input the first monitoring information into a trained neural network training model, and acquire first risk information and a first risk index according to an output result of the neural network model; Compare the first risk index with a preset threshold value, and if the first risk index is greater than or equal to the preset threshold value, obtain second monitoring information with a second granularity in a second period, wherein the second granularity is smaller than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

[0006] In conjunction with the first aspect, in some embodiments, the method further comprises: Compare the first risk index with a preset threshold value, and if the first risk index is less than the preset threshold value, obtain second monitoring information with a second granularity in a second period, wherein the second granularity is greater than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

[0007] In conjunction with the first aspect, in some embodiments, the method further comprises: The second risk index is compared with a preset threshold value. If the second risk index is greater than or equal to the preset threshold value, second monitoring information with a third granularity is obtained in a third period, wherein the third granularity is smaller than the second granularity.

[0008] In conjunction with the first aspect, in some embodiments, a training set is obtained, and a neural network training model is trained using the training set, and the neural network training model is used to output a predicted risk index and predicted risk information based on the input monitoring information and the input risk index, including: Acquire multiple monitoring information, multiple risk indexes and multiple risk information, wherein one monitoring information, one risk index and one risk information form a fixed training group, wherein the multiple monitoring information have different information granularities; Import multiple training groups into the neural network training model and perform iterative calculations. When the calculation process meets the preset conditions, stop the calculation and output the trained neural network training model.

[0009] In combination with the first aspect, in some implementations, multiple training groups are imported into a neural network training model and iterative calculations are performed, and when the calculation process meets a preset condition, the calculation is stopped and the trained neural network training model is output, including: Determine the components of monitoring information in each training group, the components including at least three monitoring data; The risk index and risk information in the same group are merged and formed into a single vector; The three monitoring data are used as three different dimensions to form the feature vector of risk information in the same training group and iteratively calculated; The three monitoring data are taken as three different dimensions to form the feature vector of the risk index in the same training group and are calculated iteratively.

[0010] In combination with the first aspect, in some implementations, multiple training groups are imported into a neural network training model and iterative calculations are performed, and when the calculation process meets a preset condition, the calculation is stopped and the trained neural network training model is output, including: Import monitoring information into the neural network training model and obtain the risk information and risk index output by the neural network model; The risk information and risk index output by the neural network model are compared with the risk information and risk index of the same training group of the monitoring information imported into the network training model, and the error value is obtained; When the error value is less than the preset error value, the training is stopped and the trained neural network training model is output.

[0011] In conjunction with the first aspect, in some embodiments, the risk information and risk index output by the neural network model are compared with the risk information and risk index of the same training group of monitoring information imported into the network training model, and an error value is obtained, including: Obtain the weights corresponding to the risk information and the risk index, where the weight corresponding to the risk information is a, and the weight corresponding to the risk index is b; Determine a first conditional value based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network training model; Determine a second conditional value based on the risk index output by the neural network model and the risk index of the same training group of monitoring information imported into the network training model; The error value is determined based on the first condition value and the second condition value, and satisfies:

[0012] in, is the error value, is the first condition value, is the second condition value.

[0013] In combination with the first aspect, in some implementations, the first condition value is determined based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network training model, satisfying: when > hour,

[0014] when < hour,

[0015] in, is the corresponding value of the risk information output by the neural network model, It is the corresponding value of the risk information of the same training group of the monitoring information imported into the network training model.

[0016] In combination with the first aspect, in some embodiments, the second condition value is determined based on the risk index output by the neural network model and the risk index of the same training group of monitoring information imported into the network training model, satisfying: when > hour,

[0017] when < hour,

[0018] in, is the corresponding value of the risk index output by the neural network model, It is the corresponding value of the risk index of the same training group of monitoring information imported into the network training model.

[0019] In a second aspect, an embodiment of the present application proposes a data processing system based on an intelligent prediction model for safety risks of power generation enterprises. The data processing system based on an intelligent prediction model for safety risks of power generation enterprises includes a controller, and the system is configured as follows: Obtaining a training set, and using the training set to train a neural network training model, the neural network training model is used to input monitoring information and input risk index, and output a predicted risk index and predicted risk information; Acquire first monitoring information with a first information granularity in a first monitoring period, input the first monitoring information into a trained neural network training model, and acquire first risk information and a first risk index according to an output result of the neural network model; Compare the first risk index with a preset threshold value, and if the first risk index is greater than or equal to the preset threshold value, obtain second monitoring information with a second granularity in a second period, wherein the second granularity is smaller than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

[0020] In conjunction with the second aspect, in some embodiments, the system is configured as follows: Compare the first risk index with a preset threshold value, and if the first risk index is less than the preset threshold value, obtain second monitoring information with a second granularity in a second period, wherein the second granularity is greater than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

[0021] In conjunction with the first aspect, in some embodiments, the method further comprises: The second risk index is compared with a preset threshold value. If the second risk index is greater than or equal to the preset threshold value, second monitoring information with a third granularity is obtained in a third period, wherein the third granularity is smaller than the second granularity.

[0022] In conjunction with the second aspect, in some embodiments, the system is configured as follows: A training set is obtained, and the training set is used to train a neural network training model. The neural network training model is used to input monitoring information and input risk index, and output predicted risk index and predicted risk information, including: Acquire multiple monitoring information, multiple risk indexes and multiple risk information, wherein one monitoring information, one risk index and one risk information form a fixed training group, wherein the multiple monitoring information have different information granularities; Import multiple training groups into the neural network training model and perform iterative calculations. When the calculation process meets the preset conditions, stop the calculation and output the trained neural network training model.

[0023] In conjunction with the second aspect, in some embodiments, the system is configured as follows: Import multiple training groups into the neural network training model and perform iterative calculations. When the calculation process meets the preset conditions, stop the calculation and output the trained neural network training model, including: Determine the components of monitoring information in each training group, the components including at least three monitoring data; The risk index and risk information in the same group are merged and formed into a single vector; The three monitoring data are used as three different dimensions to form the feature vector of risk information in the same training group and iteratively calculated; The three monitoring data are taken as three different dimensions to form the feature vector of the risk index in the same training group and are calculated iteratively.

[0024] In conjunction with the second aspect, in some embodiments, the system is configured as follows: Import multiple training groups into the neural network training model and perform iterative calculations. When the calculation process meets the preset conditions, stop the calculation and output the trained neural network training model, including: Import monitoring information into the neural network training model and obtain the risk information and risk index output by the neural network model; The risk information and risk index output by the neural network model are compared with the risk information and risk index of the same training group of the monitoring information imported into the network training model, and the error value is obtained; When the error value is less than the preset error value, the training is stopped and the trained neural network training model is output.

[0025] In conjunction with the second aspect, in some embodiments, the system is configured as follows: The risk information and risk index output by the neural network model are compared with the risk information and risk index of the same training group of the monitoring information imported into the network training model, and the error value is obtained, including: Obtain the weights corresponding to the risk information and the risk index, where the weight corresponding to the risk information is a, and the weight corresponding to the risk index is b; Determine a first conditional value based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network training model; Determine a second conditional value based on the risk index output by the neural network model and the risk index of the same training group of monitoring information imported into the network training model; The error value is determined based on the first condition value and the second condition value, and satisfies:

[0026] in, is the error value, is the first condition value, is the second condition value.

[0027] In conjunction with the second aspect, in some embodiments, the system is configured as follows: The first condition value is determined based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network training model, which satisfies: when > hour,

[0028] when < hour,

[0029] in, is the corresponding value of the risk information output by the neural network model, It is the corresponding value of the risk information of the same training group of the monitoring information imported into the network training model.

[0030] In combination with the first aspect, in some embodiments, the second condition value is determined based on the risk index output by the neural network model and the risk index of the same training group of monitoring information imported into the network training model, satisfying: when > hour,

[0031] when < hour,

[0032] in, is the corresponding value of the risk index output by the neural network model, It is the corresponding value of the risk index of the same training group of monitoring information imported into the network training model.

[0033] A third aspect of an embodiment of the present invention provides an electronic device, the electronic device comprising: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method proposed in the first aspect of the embodiment of the present invention.

[0034] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided in the first aspect of the embodiment of the present invention.

[0035] In summary, the above method and device have the following technical effects: The data processing method based on the intelligent prediction model of safety risk of power generation enterprises proposed in the embodiment of the present application, in the first monitoring period, the first monitoring information with the first information granularity is obtained, and the first monitoring information is input into the trained neural network training model, and the first risk information and the first risk index are obtained according to the output result of the neural network model, and then the first risk index is compared with the preset threshold value, and if the first risk index is greater than or equal to the preset threshold value, the second monitoring information with the second information granularity is obtained in the second period, wherein the second granularity is smaller than the first granularity, and finally, the second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model. The data processing method based on the intelligent prediction model of safety risk of power generation enterprises proposed in the embodiment of the present application, the current risk state is judged by the risk index output by the neural network training model, and when the risk state is high, in the subsequent period, the monitoring data with a smaller information granularity scale, that is, more refined monitoring data, is obtained to predict the actual risk, so that the prediction result of the second period is also more refined, which is convenient for risk investigation and control. When the risk state is low, the granularity scale corresponding to the detection data is small, which saves the detection cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A flowchart of a data processing method based on an intelligent prediction model for safety risks of power generation enterprises proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] The embodiment of the present application proposes a data processing method based on an intelligent prediction model of safety risks of power generation enterprises, which is applicable to a data processing system based on an intelligent prediction model of safety risks of power generation enterprises. The data processing system based on the intelligent prediction model of safety risks of power generation enterprises includes a controller, and the method is applicable to the controller. Figure 1 , the method comprises the following steps: S101: Obtain a training set, and use the training set to train a neural network training model, where the neural network training model is used to input monitoring information and input risk index, and output a predicted risk index and predicted risk information.

[0039] Understandably, firstly, a training set containing a large amount of historical monitoring information and risk index is constructed. These data cover all aspects of the daily operation of power generation enterprises and provide rich samples for model training. The neural network training model can be trained in combination with this training set.

[0040] During the training process, the model gradually formed the ability to effectively predict the safety risks of power generation enterprises, such as risk type, possible impact range, etc., by learning the complex relationship between the input monitoring information (such as equipment operating status, ambient temperature, humidity, etc.) and the risk index. This information is crucial for power generation enterprises to formulate effective risk prevention measures.

[0041] Specifically, in this embodiment, multiple monitoring information, multiple risk indexes, and multiple risk information can be obtained. One monitoring information, one risk index, and one risk information form a fixed training group, wherein the multiple monitoring information have different information granularities, which facilitates subsequent prediction by combining data of different information granularities. Afterwards, the multiple training groups can be imported into the neural network training model and iteratively calculated. When the calculation process meets the preset conditions, the calculation is stopped and the trained neural network training model is output.

[0042] It is understandable that the training process of the neural network has various methods according to different neural network structures. In this embodiment, the following methods S201-S204 are taken as examples: S201: Determine components of monitoring information in each training group, where the components include at least three monitoring data.

[0043] It can be understood that the components of the detection information include various information such as equipment operating status, ambient temperature, humidity, etc. The various information may correspond to situations with different information granularity. For example, the operating status of the equipment, as the information granularity decreases, may be the operating status of a production line, the operating status of a workstation, or the operating status of a device.

[0044] The risk index and risk information in the same group are fused and formed into a single vector.

[0045] The three monitoring data are taken as three different dimensions to form the feature vector of risk information in the same training group and are calculated iteratively.

[0046] The three monitoring data are taken as three different dimensions to form the feature vector of the risk index in the same training group and are calculated iteratively.

[0047] Through the above methods, multi-dimensional monitoring data can be effectively converted into risk index and risk information feature vectors, thereby achieving more accurate risk prediction in the neural network model. This multi-dimensional feature representation not only considers the correlation between different monitoring data, but also makes full use of the richness brought by different information granularity, so that the model can capture more detailed and comprehensive risk characteristics.

[0048] Furthermore, in the iterative calculation process of the feature vector, various optimization algorithms can be used to continuously adjust the weights and biases in order to achieve the best prediction effect. These optimization algorithms include but are not limited to the gradient descent method, Newton method, quasi-Newton method, conjugate gradient method, etc., which can help find the parameter combination that minimizes the loss function.

[0049] In addition, in order to further improve the generalization ability and robustness of the model, regularization terms, dropout and other strategies can be introduced during the training process to prevent overfitting. Regularization terms reduce the complexity of the model by constraining parameters, while dropout avoids the model's excessive reliance on specific features by randomly discarding some neurons.

[0050] In summary, through the fusion of multi-dimensional monitoring data, the construction of feature vectors and the application of optimization algorithms, an efficient and accurate intelligent prediction model for safety risks of power generation enterprises can be constructed. This model can not only evaluate the safety risk level of enterprises in real time and accurately, but also provide enterprises with effective risk prevention and control suggestions, thereby ensuring the safe production and stable operation of enterprises.

[0051] As for how to determine the stop node of neural network training, illustratively, a monitoring information can be imported into the neural network training model and the risk information and risk index of the output of the neural network model can be obtained. Then, the risk information and risk index of the output of the neural network model are compared with the risk information and risk index of the same training group of the monitoring information imported into the network training model, and the error value is obtained. Finally, when the error value is less than the preset error value, the training is stopped and the trained neural network training model is output.

[0052] Specifically, in this embodiment, the weights corresponding to the risk information and the risk index may be obtained, wherein the weight corresponding to the risk information is a, and the weight corresponding to the risk index is b; The first condition value is determined based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network training model, wherein: when > hour,

[0053] when < hour,

[0054] is the corresponding value of the risk information output by the neural network model, It is the corresponding value of the risk information of the same training group of the monitoring information imported into the network training model.

[0055] The second conditional value is determined based on the risk index output by the neural network model and the risk index of the same training group of monitoring information imported into the network training model, wherein: when > hour,

[0056] when < hour,

[0057] in, is the corresponding value of the risk index output by the neural network model, It is the corresponding value of the risk index of the same training group of monitoring information imported into the network training model.

[0058] Finally, the error value is determined based on the first condition value and the second condition value, and satisfies:

[0059] in, is the error value, is the first condition value, is the second condition value.

[0060] In some other embodiments, in addition to the above-mentioned method of determining the stop node of neural network training by comparing error values, cross-validation technology can also be used to further optimize the model training process. Cross-validation is a method for evaluating the performance of a statistical model by dividing a data set into multiple parts, using one part as a training set and the rest as a test set in turn to evaluate the model's predictive ability. This method can not only be used to select the best model parameters, but also effectively avoid overfitting and ensure that the model has good generalization ability.

[0061] In the context of neural network training, K-Fold Cross-Validation can be used. Specifically, the monitoring data set is divided into K subsets, and K-1 subsets are selected as training sets each time, and the remaining subset is used as the validation set. In this way, K training and validation can be performed, and an error value will be obtained for each validation. Finally, the average of these K error values ​​can be calculated as an evaluation indicator of model performance. When this average error value does not change much for many consecutive times or reaches a preset threshold, it can be considered that the model has been trained well enough and training can be stopped.

[0062] In addition, you can also consider using the Early Stopping method to terminate the training process early. The Early Stopping method monitors the error value on the validation set during the training process. If the error value does not drop significantly in multiple consecutive training cycles (epochs), it can be considered that the model has fallen into the risk of overfitting, and training should be stopped at this time. This method can not only save computing resources, but also effectively prevent the model from performing too well on the training set and performing poorly on the test set.

[0063] S102: Acquire first monitoring information with a first information granularity in a first monitoring period, input the first monitoring information into a trained neural network training model, and acquire first risk information and a first risk index according to an output result of the neural network model.

[0064] It can be understood that after obtaining the trained neural network training model, in the first monitoring cycle, that is, the actual operation cycle, the monitoring data can be imported into the trained neural network training model. At this time, the first monitoring information corresponding to the first cycle can be obtained, and its information granularity can be the first granularity, for example, the information is the overall operation of a machine in a factory. Based on the obtained data and importing the data into the neural network training model, the first risk information and the first risk index predicted by the data after the first detection cycle can be obtained. The first risk index can be used to characterize the risk situation in the first operation cycle, and the first risk information can be used to characterize the specific characteristics of the risk, such as location, region, etc.

[0065] S103: Compare the first risk index with a preset threshold value. If the first risk index is greater than or equal to the preset threshold value, obtain second monitoring information with a second granularity in a second period, wherein the second granularity is smaller than the first granularity.

[0066] It is understandable that after obtaining the first risk index, by measuring the first risk index, a rough judgment can be made on the current risk status. In this embodiment, the subsequent detection method can be determined based on the relationship between the first risk index and the preset threshold. Specifically, in this embodiment, a preset threshold can be set. For example, when the risk index unit is a percentage, the threshold can be set to 50%. When the risk index is higher than 50%, it can be judged that it is in a high risk state at this time.

[0067] It can be understood that after determining that the risk status is high, the granularity of obtaining detection information is correspondingly reduced in subsequent detection cycles. For example, in the second cycle, the acquisition of detection data can be refined into the detection process of the production line or independent production equipment. In this way, in the results predicted based on the detection data of the same information granularity, the information granularity of the second risk information is also smaller than that of the first risk information, that is, the second risk information is more detailed than the first risk information. Therefore, in the process of troubleshooting risks based on the prediction results, risks and related issues can be more easily discovered.

[0068] Of course, the first risk index is compared with the preset threshold. If the first risk index is less than the preset threshold, second monitoring information with a second information granularity is obtained in the second period, wherein the second granularity is greater than the first granularity. Then, the second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

[0069] It can be understood that when the first risk index is less than the preset threshold, it can be set to a safe mode. At this time, from the perspective of monitoring cost, monitoring information with larger granularity can be obtained in the second monitoring cycle. For example, the detection data at this time can be the detection data of the entire factory area.

[0070] S104: Input the second monitoring information into the trained neural network training model, and obtain the second risk information and the second risk index according to the output result of the neural network model.

[0071] S105: Compare the second risk index with a preset threshold value. If the second risk index is greater than or equal to the preset threshold value, obtain second monitoring information with a third granularity in a third period, wherein the third granularity is smaller than the second granularity.

[0072] It can be understood that in order to enable the system to run continuously, the data obtained in the second monitoring cycle also becomes the measurement standard for the acquisition granularity when acquiring monitoring data in the third cycle, that is, the data of the third monitoring cycle also affects the fourth monitoring cycle. In this way, the system can run continuously.

[0073] The data processing method based on the intelligent prediction model of safety risk of power generation enterprises proposed in the embodiment of the present application, in the first monitoring period, the first monitoring information with the first information granularity is obtained, and the first monitoring information is input into the trained neural network training model, and the first risk information and the first risk index are obtained according to the output result of the neural network model, and then the first risk index is compared with the preset threshold value, and if the first risk index is greater than or equal to the preset threshold value, the second monitoring information with the second information granularity is obtained in the second period, wherein the second granularity is smaller than the first granularity, and finally, the second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model. The data processing method based on the intelligent prediction model of safety risk of power generation enterprises proposed in the embodiment of the present application, the current risk state is judged by the risk index output by the neural network training model, and when the risk state is high, in the subsequent period, the monitoring data with a smaller information granularity scale, that is, more refined monitoring data, is obtained to predict the actual risk, so that the prediction result of the second period is also more refined, which is convenient for risk investigation and control. When the risk state is low, the granularity scale corresponding to the detection data is small, which saves the detection cost.

[0074] In a second aspect, an embodiment of the present application proposes a data processing system based on an intelligent prediction model for safety risks of power generation enterprises. The data processing system based on an intelligent prediction model for safety risks of power generation enterprises includes a controller, and the system is configured as follows: Obtaining a training set, and using the training set to train a neural network training model, the neural network training model is used to input monitoring information and input risk index, and output a predicted risk index and predicted risk information; Acquire first monitoring information with a first information granularity in a first monitoring period, input the first monitoring information into a trained neural network training model, and acquire first risk information and a first risk index according to an output result of the neural network model; Compare the first risk index with a preset threshold value, and if the first risk index is greater than or equal to the preset threshold value, obtain second monitoring information with a second granularity in a second period, wherein the second granularity is smaller than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

[0075] In conjunction with the second aspect, in some embodiments, the system is configured as follows: Compare the first risk index with a preset threshold value, and if the first risk index is less than the preset threshold value, obtain second monitoring information with a second granularity in a second period, wherein the second granularity is greater than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

[0076] In conjunction with the first aspect, in some embodiments, the method further comprises: The second risk index is compared with a preset threshold value. If the second risk index is greater than or equal to the preset threshold value, second monitoring information with a third granularity is obtained in a third period, wherein the third granularity is smaller than the second granularity.

[0077] In conjunction with the second aspect, in some embodiments, the system is configured as follows: A training set is obtained, and the training set is used to train a neural network training model. The neural network training model is used to input monitoring information and input risk index, and output predicted risk index and predicted risk information, including: Acquire multiple monitoring information, multiple risk indexes and multiple risk information, wherein one monitoring information, one risk index and one risk information form a fixed training group, wherein the multiple monitoring information have different information granularities; Import multiple training groups into the neural network training model and perform iterative calculations. When the calculation process meets the preset conditions, stop the calculation and output the trained neural network training model.

[0078] In conjunction with the second aspect, in some embodiments, the system is configured as follows: Import multiple training groups into the neural network training model and perform iterative calculations. When the calculation process meets the preset conditions, stop the calculation and output the trained neural network training model, including: Determine the components of monitoring information in each training group, the components including at least three monitoring data; The risk index and risk information in the same group are merged and formed into a single vector; The three monitoring data are used as three different dimensions to form the feature vector of risk information in the same training group and iteratively calculated; The three monitoring data are taken as three different dimensions to form the feature vector of the risk index in the same training group and are calculated iteratively.

[0079] In conjunction with the second aspect, in some embodiments, the system is configured as follows: Import multiple training groups into the neural network training model and perform iterative calculations. When the calculation process meets the preset conditions, stop the calculation and output the trained neural network training model, including: Import monitoring information into the neural network training model and obtain the risk information and risk index output by the neural network model; The risk information and risk index output by the neural network model are compared with the risk information and risk index of the same training group of the monitoring information imported into the network training model, and the error value is obtained; When the error value is less than the preset error value, the training is stopped and the trained neural network training model is output.

[0080] In conjunction with the second aspect, in some embodiments, the system is configured as follows: The risk information and risk index output by the neural network model are compared with the risk information and risk index of the same training group of the monitoring information imported into the network training model, and the error value is obtained, including: Obtain the weights corresponding to the risk information and the risk index, where the weight corresponding to the risk information is a, and the weight corresponding to the risk index is b; Determine a first conditional value based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network training model; Determine a second conditional value based on the risk index output by the neural network model and the risk index of the same training group of monitoring information imported into the network training model; The error value is determined based on the first condition value and the second condition value, and satisfies:

[0081] in, is the error value, is the first condition value, is the second condition value.

[0082] In conjunction with the second aspect, in some embodiments, the system is configured as follows: The first condition value is determined based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network training model, which satisfies: when > hour,

[0083] when < hour,

[0084] in, is the corresponding value of the risk information output by the neural network model, It is the corresponding value of the risk information of the same training group of the monitoring information imported into the network training model.

[0085] In combination with the first aspect, in some embodiments, the second condition value is determined based on the risk index output by the neural network model and the risk index of the same training group of monitoring information imported into the network training model, satisfying: when > hour,

[0086] when < hour,

[0087] in, is the corresponding value of the risk index output by the neural network model, It is the corresponding value of the risk index of the same training group of monitoring information imported into the network training model.

[0088] The embodiment of the present application proposes a data processing system based on the intelligent prediction model of safety risks of power generation enterprises. In the first monitoring period, first monitoring information with a first information granularity is obtained, and the first monitoring information is input into the trained neural network training model. According to the output result of the neural network model, the first risk information and the first risk index are obtained. Then, the first risk index is compared with the preset threshold. If the first risk index is greater than or equal to the preset threshold, the second monitoring information with a second information granularity is obtained in the second period, wherein the second granularity is smaller than the first granularity. Finally, the second monitoring information is input into the trained neural network training model. According to the output result of the neural network model, the second risk information and the second risk index are obtained. The embodiment of the present application proposes a data processing system based on the intelligent prediction model of safety risks of power generation enterprises. The current risk status is judged by the risk index output by the neural network training model. When the risk status is high, in the subsequent period, monitoring data with a smaller information granularity, that is, more refined monitoring data, is obtained to predict the actual risk, so that the prediction result of the second period is also more refined, which is convenient for risk investigation and control.

[0089] Based on the same inventive concept, an embodiment of the present application further proposes an electronic device, the electronic device comprising: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to an embodiment of the present application.

[0090] In addition, to achieve the above-mentioned purpose, an embodiment of the present application also proposes a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the data processing method based on the intelligent prediction model of safety risks of power generation enterprises in the embodiment of the present application.

[0091] The following is a detailed introduction to the various components of electronic equipment: The processor is the control center of the electronic device, which can be a processor or a general term for multiple processing elements. For example, the processor is one or more central processing units (CPUs), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention, such as one or more microprocessors (digital signal processors, DSPs), or one or more field programmable gate arrays (FPGAs).

[0092] Optionally, the processor can perform various functions of the electronic device by running or executing a software program stored in the memory, and calling data stored in the memory.

[0093] The memory is used to store the software program for executing the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can refer to the above method embodiment and will not be repeated here.

[0094] Optionally, the memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may be integrated with the processor, or may exist independently and be coupled to the processor through an interface circuit of the electronic device, which is not specifically limited in the embodiments of the present invention.

[0095] A transceiver is used to communicate with a network device or a terminal device.

[0096] Optionally, the transceiver may include a receiver and a transmitter, wherein the receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.

[0097] Optionally, the transceiver may be integrated with the processor, or may exist independently and be coupled to the processor via an interface circuit of the router, which is not specifically limited in the embodiment of the present invention.

[0098] In addition, the technical effects of the electronic device can refer to the technical effects of the data transmission method in the above method embodiment, which will not be repeated here.

[0099] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0100] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0101] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments 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 or computer programs. When a computer instruction or computer program is loaded or executed on a computer, a process or function according to an embodiment of the present invention 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. Computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0102] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0103] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0104] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0105] Those skilled 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 the present invention.

Claims

1. A data processing method based on an intelligent prediction model for safety risks of power generation enterprises, characterized in that: The invention is applicable to a data processing system based on a power generation enterprise safety risk intelligent prediction model, wherein the data processing system based on the power generation enterprise safety risk intelligent prediction model comprises a controller, and the method is applicable to the controller, comprising: Obtaining a training set, and using the training set to train a neural network training model, wherein the neural network training model is used to output a predicted risk index and predicted risk information based on input monitoring information and input risk index; Acquire first monitoring information with a first information granularity in a first monitoring period, input the first monitoring information into the trained neural network training model, and acquire first risk information and a first risk index according to an output result of the neural network model; Comparing the first risk index with a preset threshold, if the first risk index is greater than or equal to the preset threshold, acquiring second monitoring information with a second granularity in a second period, wherein the second granularity is smaller than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

2. According to claim 1, a data processing method based on the intelligent prediction model of power generation enterprise safety risk is characterized in that: The method further comprises: Comparing the first risk index with a preset threshold, if the first risk index is less than the preset threshold, acquiring second monitoring information with a second granularity in a second period, wherein the second granularity is greater than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

3. A data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to claim 2, characterized in that: The method further comprises: The second risk index is compared with a preset threshold. If the second risk index is greater than or equal to the preset threshold, second monitoring information with a third granularity is obtained in a third period, wherein the third granularity is smaller than the second granularity.

4. A data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to claim 3, characterized in that: A training set is obtained, and a neural network training model is trained using the training set. The neural network training model is used to output a predicted risk index and predicted risk information based on the input monitoring information and the input risk index, including: Acquire a plurality of the monitoring information, a plurality of the risk indexes and a plurality of the risk information, wherein one piece of the monitoring information, one piece of the risk index and one piece of the risk information form a fixed training group, wherein the plurality of the monitoring information have different information granularities; The plurality of training groups are imported into the neural network training model and iterative calculation is performed. When the calculation process meets the preset conditions, the calculation is stopped and the trained neural network training model is output.

5. A data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to claim 4, characterized in that: Importing the plurality of training groups into the neural network training model and performing iterative calculations, stopping the calculations when the calculation process meets a preset condition and outputting the trained neural network training model, including: Determine components of the monitoring information in each of the training groups, wherein the components include at least three monitoring data; Merging the risk index and the risk information in the same group into a single vector; The three monitoring data are used as three different dimensions to form a feature vector of the risk information in the same training group and perform iterative calculation; The three monitoring data are used as three different dimensions to form the characteristic vector of the risk index in the same training group and perform iterative calculation.

6. A data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to claim 5, characterized in that: Importing the plurality of training groups into the neural network training model and performing iterative calculations, stopping the calculations when the calculation process meets a preset condition and outputting the trained neural network training model, including: Importing the monitoring information into the neural network training model and obtaining the risk information and the risk index output by the neural network model; The risk information and the risk index output by the neural network model are compared with the risk information and the risk index of the same training group of the monitoring information imported into the network training model, and an error value is obtained; When the error value is less than the preset error value, the training is stopped and the trained neural network training model is output.

7. A data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to claim 6, characterized in that: The risk information and the risk index output by the neural network model are compared with the risk information and the risk index of the same training group of the monitoring information imported into the network training model, and an error value is obtained, including: Obtaining weights corresponding to the risk information and the risk index, wherein the weight corresponding to the risk information is a, and the weight corresponding to the risk index is b; Determine a first conditional value based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network trained model; Determine a second conditional value based on the risk index output by the neural network model and the risk index of the same training group of the monitoring information imported into the network trained model; The error value is determined based on the first condition value and the second condition value, and satisfies: in, is the error value, is the first condition value, is the second condition value.

8. A data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to claim 7, characterized in that: The first condition value is determined based on the risk information output by the neural network model and the risk information of the same training group of the monitoring information imported into the network trained model, which satisfies: when > hour, when < hour, in, is the corresponding value of the risk information output by the neural network model, The corresponding values ​​of the risk information of the same training group as the monitoring information imported into the network training model.

9. The data processing method based on the intelligent prediction model of safety risks of power generation enterprises according to claim 7 is characterized in that: The second condition value is determined based on the risk index output by the neural network model and the risk index of the same training group of the monitoring information imported into the network trained model, satisfying: when > hour, when < hour, in, is the corresponding value of the risk index output by the neural network model, The corresponding value of the risk index of the same training group as the monitoring information imported into the network training model.

10. A data processing system based on an intelligent prediction model for safety risks of power generation enterprises, characterized in that: The data processing system based on the intelligent prediction model of safety risks of power generation enterprises includes a controller, which is configured as follows: Obtaining a training set, and using the training set to train a neural network training model, wherein the neural network training model is used to output a predicted risk index and predicted risk information based on input monitoring information and input risk index; Acquire first monitoring information with a first information granularity in a first monitoring period, input the first monitoring information into the trained neural network training model, and acquire first risk information and a first risk index according to an output result of the neural network model; Comparing the first risk index with a preset threshold, if the first risk index is greater than or equal to the preset threshold, acquiring second monitoring information with a second granularity in a second period, wherein the second granularity is smaller than the first granularity; The second monitoring information is input into the trained neural network training model, and the second risk information and the second risk index are obtained according to the output result of the neural network model.

Images