CN117391450A

CN117391450A - FTA-ANN-based risk evaluation method and device for atmospheric storage tank area and electronic equipment

Info

Publication number: CN117391450A
Application number: CN202311445302.5A
Authority: CN
Inventors: 周宁; 赵鹏飞; 李雪; 黄为波; 黄维秋; 赵会军; 袁雄军; 冯胜; 刘为奥
Original assignee: Changzhou University
Current assignee: Changzhou University
Priority date: 2023-11-01
Filing date: 2023-11-01
Publication date: 2024-01-12

Abstract

The application relates to the field of safety risk evaluation of storage tank areas in chemical industry parks, in particular to a risk evaluation method, a risk evaluation device and electronic equipment for an atmospheric storage tank area based on FTA-ANN, which can ensure safe and stable operation of the atmospheric storage tank area to a certain extent and effectively prevent major accidents. The atmospheric storage tank region risk evaluation method based on FTA-ANN comprises the following steps: building a fault tree model, identifying dangerous sources in the area of the normal pressure storage tank, and building to obtain the fault tree model; establishing an FTA-ANN model, mapping the fault tree model into an ANN, and performing multiple training through a training model to obtain an FTA-ANN model; and performing performance test and risk assessment, predicting based on the FTA-ANN model to obtain a prediction result, and comparing the prediction result with the FTA prediction result to realize risk assessment on the normal pressure storage tank area.

Description

FTA-ANN-based risk evaluation method and device for atmospheric storage tank area and electronic equipment

Technical Field

The application relates to the field of safety risk evaluation of storage tanks in chemical industry parks, in particular to a method, a device and electronic equipment for evaluating risks of an atmospheric storage tank region based on FTA-ANN.

Background

Fault tree analysis (Fault Tree Analysis, FTA for short) is a tool for system fault analysis and reliability assessment. It helps us understand faults that may occur in the system and their potential impact by graphically representing the logical relationships between the potential fault causes. By combining basic events and logic gates according to certain rules, a fault tree model can be constructed. In the fault tree analysis process, the probability theory, the Boolean algebra and other methods can be used for quantitatively analyzing the fault tree, and the probability of occurrence of the top event or the reliability index of the system can be calculated. The method can help us identify potential risks and fragile links in the system, find out a main fault path causing the occurrence of the top event, and provide basis for system design improvement and risk control.

An artificial neural network (Artificial Neural Network, ANN for short) is a computational model that simulates the principle of operation of the human nervous system. It consists of a large number of interconnected artificial neurons (also called nodes or cells) that transmit signals by weighting and process them by an activation function.

The chemical industry is the pillar industry of national economy, because chemical products generally have characteristics such as inflammable, explosive, and the like, and the atmospheric storage tank is the main storage means of chemical products again, and the liquid medium in the atmospheric storage tank often possesses physical properties such as inflammable and explosive, once the atmospheric storage tank breaks or leaks, will take place conflagration or explosion after meetting open fire, and once the atmospheric storage tank takes place the accident and often can influence other storage tanks in same cofferdam, the storage tank accident can cause other storage tank accidents, causes the accident to upgrade, causes more serious influence, consequently carries out the safety assessment to the storage tank in chemical industry garden very much necessary.

At present, in order to ensure safe and stable operation of the normal pressure storage tank area, major accidents are effectively prevented, and prediction of the normal pressure storage tank accidents is very necessary.

Disclosure of Invention

In order to ensure safe and stable operation of the normal pressure storage tank and effectively prevent major accidents, prediction of the normal pressure storage tank accidents is very necessary, and the application provides a normal pressure storage tank regional risk evaluation method, device and electronic equipment based on FTA-ANN.

Embodiments of the present application are implemented as follows:

in a first aspect, the present application provides an atmospheric storage tank area risk assessment method based on FTA-ANN, including:

building a fault tree model, identifying a dangerous source of the normal pressure storage tank, and building to obtain the fault tree model;

establishing an FTA-ANN model, mapping the fault tree model into an ANN, and performing multiple training through a training model to obtain an FTA-ANN model;

and performing performance test and risk assessment, predicting based on the FTA-ANN model to obtain a prediction result, and comparing the prediction result with the FTA prediction result to realize risk assessment on the normal pressure storage tank.

In one possible implementation, most of the cause of a tank fire or explosion is due to tank leakage, and a tank leakage event is identified as a top event in the fault tree model.

In one possible implementation, the mapping rules for mapping the fault tree model into an ANN include:

the basic events are mapped to input nodes, the intermediate events are mapped to synaptic weights, the logic gates are mapped to transfer functions and synaptic weights, and the top events are mapped to output nodes.

the basic events are mapped to input nodes, the intermediate events are mapped to hidden nodes, the logic gates are mapped to transfer functions and synaptic weights, and the top events are mapped to output nodes.

In one possible implementation, the training model includes one input layer, two hidden layers, and one output layer.

In one possible implementation manner, the number of nodes of the two hidden layers is selected according to a method of hiding the number of neurons and combining the characteristics of the fault tree model, and the selection rule includes:

the total number of hidden neurons should be between the number of input layer neurons and the number of output layer neurons;

the number of hidden neurons should be the sum of 2/3 of the number of input layer neurons and the number of output layer neurons;

if the two rules are satisfied, the number of the neurons of the first hidden layer should be equivalent to the number of the intermediate events directly connected with the basic events in the fault tree model, and if the second rule is not satisfied, the maximum number of the neurons of the first hidden layer should be the maximum total number of the hidden neurons allowed minus the number of the intermediate events directly connected with TE (top-level events);

the number of neurons of the second hidden layer should be the same as the number of intermediate events in the fault tree model that are directly connected to the TE.

In one possible implementation, the training model is trained by a back propagation algorithm, an optimizer adopts Adam algorithm, uses mean square error as a loss function, and uses a method of small batch random gradient descent to optimize in the training process, the data is divided into small batches to train, and the maximum iteration times and an MSE threshold value in the iteration process are set to monitor the model performance. When the MSE has not improved for 10 consecutive iterations, the training process will stop.

In one possible implementation manner, in the performance test and risk assessment, the application performs sensitivity analysis on input variables to judge the influence of each accident factor on the failure of the top-level event, and the calculation formula is as follows:

wherein CRIT _i The critical importance of the input variable i, i.e., the degree of influence of the input variable i on the final output, P (TOP) is the system failure probability, and P (i) is the component i failure probability.

In a second aspect, the present application provides an atmospheric storage tank area risk assessment device based on FTA-ANN, comprising:

the model building module is used for building a fault tree model, identifying dangerous sources in the area of the normal pressure storage tank and building to obtain the fault tree model;

the model mapping and training module is used for establishing an FTA-ANN model, mapping the fault tree model into an ANN, and performing multiple training through the training model to establish an FTA-ANN model;

and the model test and evaluation module is used for performance test and risk evaluation, predicting based on the FTA-ANN model to obtain a predicted result, and comparing the predicted result with the FTA predicted result to realize risk evaluation on the normal pressure storage tank area.

In a third aspect, the present application provides an electronic device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when calling and executing the computer program from the memory, implements the steps of any of the methods described above.

The technical scheme provided by the application can at least achieve the following beneficial effects:

according to the atmospheric storage tank risk evaluation method, the atmospheric storage tank risk evaluation device and the atmospheric storage tank risk evaluation electronic equipment based on the FTA-ANN, aiming at the risk of fire explosion of the storage tank, a fault tree model is built and mapped into an ANN, and training is carried out for a plurality of times to build the FTA-ANN model, wherein the model considers the dynamics in accident analysis, and meanwhile considers the interdependence among basic events causing system faults. Compared with the traditional fault tree analysis method, the model can analyze accident reasons more accurately and rapidly and identify risk factors, so that daily safety management and accident emergency management are improved, and the safety of a chemical tank farm is ensured.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief description will be given below of the drawings that are needed in the embodiments or the prior art descriptions, it being obvious that the drawings in the following description are some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art.

FIG. 1 is a flow chart of an exemplary FTA-ANN based method for risk assessment of an atmospheric storage tank area according to an exemplary embodiment of the present application;

FIG. 2 is a schematic structural diagram of an apparatus for risk assessment of an atmospheric storage tank area based on FTA-ANN according to an exemplary embodiment of the present application;

FIG. 3 is a schematic diagram of an electronic device according to an exemplary embodiment of the present application;

FIG. 4 is a specific flow diagram of risk assessment, as shown in an exemplary embodiment of the present application;

FIG. 5 is a schematic diagram of a "tank leak" fault tree shown in an exemplary embodiment of the present application;

FIG. 6 is a schematic diagram of FT to ANN mapping as shown in an exemplary embodiment of the present application;

FIG. 7 is a schematic diagram of an ANN network of a configuration as shown in an exemplary embodiment of the present application;

FIG. 8 is a schematic diagram illustrating a comparison of DFT and ANN predictions in accordance with an exemplary embodiment of the present application;

FIG. 9 is a graph showing the degree of impact of a base event on a top event versus a top event according to an exemplary embodiment of the present application.

Reference numerals:

1. a model building module; 2. model mapping and training modules; 3. and a model test and evaluation module.

Detailed Description

For purposes of making the objects, embodiments and advantages of the present application more apparent, the exemplary embodiments of the present application will be described in detail and fully in connection with the accompanying drawings in which exemplary embodiments of the present application are shown, it being understood that the exemplary embodiments described are only some, but not all, of the examples of the present application, and it is to be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the application.

It should be noted that the brief description of the terms in the present application is only for convenience in understanding the embodiments described below, and is not intended to limit the embodiments of the present application. Unless otherwise indicated, these terms should be construed in their ordinary and customary meaning.

The terms "first," second, "" third and the like in the description and in the claims and in the above-described figures are used for distinguishing between similar or similar objects or entities and not necessarily for limiting a particular order or sequence, unless otherwise indicated. It is to be understood that the terms so used are interchangeable under appropriate circumstances.

The terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a product or apparatus that comprises a list of elements is not necessarily limited to all elements explicitly listed, but may include other elements not expressly listed or inherent to such product or apparatus.

Before explaining the method for evaluating the risk of the atmospheric storage tank area based on the FTA-ANN provided by the embodiment of the application, the application scene and the implementation environment of the embodiment of the application are described.

The chemical industry is the pillar industry of national economy, and the normal pressure storage tank is a main storage means of chemical products because the chemical products generally have the characteristics of inflammability, explosiveness and the like. Once the normal pressure storage tank is broken or leaked, fire or explosion can occur after the normal pressure storage tank is exposed to open fire. Each normal pressure storage tank in the same cofferdam is built nearby, when one storage tank is in fire or explosion, the probability of accident of the nearby normal pressure storage tank is greatly improved, and the severity of accident results is improved.

Based on the method, the device and the electronic equipment for evaluating the risk of the atmospheric storage tank based on the FTA-ANN, aiming at the risk of fire explosion of the storage tank, a fault tree model is built, mapped into an ANN and trained for multiple times to build the FTA-ANN model, and the model considers the dynamics in accident analysis and the interdependence among basic events causing system faults. Compared with traditional FTA, ANN has advantages in terms of data driving, no prior information of dependent event relationship and less dependent expert judgment. Compared with the traditional fault tree analysis method, the model can analyze accident reasons more accurately and rapidly and identify risk factors, so that daily safety management and accident emergency management are improved, and the safety of a chemical tank farm is ensured.

Next, the technical solutions of the present application and how the technical solutions of the present application solve the above technical problems will be specifically described by way of examples with reference to the accompanying drawings. Embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. It will be apparent that the described embodiments are some, but not all, of the embodiments of the present application.

FIG. 1 is a flow chart of an exemplary FTA-ANN based method for risk assessment of an area of an atmospheric storage tank.

In an exemplary embodiment, as shown in fig. 1, there is provided an atmospheric tank area risk assessment method based on FTA-ANN, and in this embodiment, the method may include the following steps:

step 100: and (3) establishing a fault tree model, identifying dangerous sources in the area of the normal pressure storage tank, and constructing to obtain the fault tree model.

Step 200: and establishing an FTA-ANN model, mapping the fault tree model into an ANN, and performing multiple training through a training model to obtain an FTA-ANN model.

Step 300: and performing performance test and risk assessment, predicting based on the FTA-ANN model to obtain a prediction result, and comparing the prediction result with the FTA prediction result to realize risk assessment on the normal pressure storage tank area in one possible implementation mode.

Fig. 4 is a specific flow chart of risk assessment according to an exemplary embodiment of the present application, and fig. 5 is a schematic diagram of a "tank leak" fault tree according to an exemplary embodiment of the present application.

In one possible implementation, as shown in fig. 4, step 100: the specific process for establishing the fault tree is as follows:

the specific process is as follows:

according to the fact that most of the reasons for a fire or explosion of a tank are due to a leakage of the tank in many tank accidents, the event of a leakage of the tank is identified as a top event in a fault tree. A fault tree is constructed as shown in fig. 5.

The basic events in the fault tree are shown in table 1 below:

TABLE 1 basic event description and failure probability thereof

Before the training is performed for a plurality of times, the method further comprises the step of effectively using fault samples and the number of the samples in the fault tree model. When the availability of the sample data is insufficient, the sample needs to be collected again; when the number of samples is insufficient, the data is expanded by a boost method, and then calculated to obtain preliminary data, and specific expanded data are shown in table 2 according to the collected data.

TABLE 2 failure probability after data expansion

FIG. 6 is a diagram illustrating FT to ANN mapping according to an exemplary embodiment of the present application.

Step 200: the specific process of establishing the FTA-ANN model is as follows:

mapping of FT to ANN 950 samples were used as training data and 50 samples were used to test the ANN according to the rules shown in fig. 6 and the flow shown in fig. 1. The failure probabilities of all 37 basic events are all used as input variables, and the top-level event is an output variable. The training model is provided with an input layer, two hidden layers and an output layer, the number of nodes of the two hidden layers is selected according to the method for hiding the number of neurons and combining the characteristics of fault trees, and the following rules are satisfied:

(1) The total number of hidden neurons should be between the number of input layer neurons and the number of output layer neurons.

(2) The number of hidden neurons should be the sum of 2/3 of the number of input layer neurons and the number of output layer neurons.

(3) If the two rules are satisfied, the number of neurons of the first hidden layer should BE equivalent to the number of IEs (intermediate events) directly connected to BE (base events) in FT (fault tree), and if the second rule is not satisfied, the number of neurons of the first hidden layer should BE the maximum allowable total number of hidden neurons minus the number of IEs directly connected to TE (top events).

(4) The number of neurons of the second hidden layer should be the same as the number of IEs in FT directly connected to TE.

Fig. 7 is a schematic diagram of an ANN network of a configuration as illustrated in an exemplary embodiment of the present application.

The fault tree diagram according to fig. 5 can be represented by 15 and 3 hidden layer neurons, respectively, and the neural network model is constructed as shown in fig. 7. A back propagation algorithm is used herein to train the neural network, where the optimizer employs Adam's algorithm, using the mean square error (Mean Squared Error, MSE) as a loss function. In the training process, a small batch random gradient descent method is used for optimization, and data are divided into small batches for training. The maximum number of iterations and the MSE threshold during the iteration are set to monitor model performance. When the MSE has not improved for 10 consecutive iterations, the training process will stop. The above method is implemented using a PyTorch neural network framework in Python.

FIG. 8 is a schematic diagram showing comparison of DFT and ANN predictions according to an exemplary embodiment of the present application, and FIG. 9 is a schematic diagram showing comparison of the degree of impact of a base event on a top event according to an exemplary embodiment of the present application.

Step 300: the specific process of performance test and risk assessment is as follows:

the developed ANNs are subjected to feasibility tests, FT is simulated by using different input data, the failure probability of the top event is predicted, and the ANN prediction set comprises 30 groups of cases of different BE failure rates and corresponding TE failure rates. The methods presented herein are compared to conventional fault tree analysis methods.

As shown in fig. 8, the result of the conventional fault tree analysis method and the model solving result provided herein are compared (wherein green is the result of the fault tree analysis method for solving the failure probability of the top event, and orange is the result of the ANN model solving), so that the matching degree of the two results is good. For the mapping performance check, the average difference between the two model results, the maximum difference and the MSE are also calculated: the average difference was 4.59389E-06, the maximum difference was 2.75278E-05, and the MSE was 1.14417E-11.

To determine the impact of each incident factor on top level event failure, a sensitivity analysis is performed on the input variables by the proposed method. Each set of data in table 2 was subjected to sensitivity calculation, then averaged, and finally the obtained data were normalized, and the obtained results are shown in fig. 9. The calculation formula is as follows:

It should be understood that, although the steps in the flowcharts relating to the above embodiments are shown in order as indicated, these steps are not necessarily performed in order as indicated. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of performing the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or steps.

Corresponding to the embodiment of the method for evaluating the risk of the atmospheric storage tank area based on the FTA-ANN, the application also provides the embodiment of the device for evaluating the risk of the atmospheric storage tank area based on the FTA-ANN by adopting the same technical conception.

Fig. 2 is a schematic structural diagram of an apparatus for risk assessment of an atmospheric storage tank area based on FTA-ANN according to an exemplary embodiment of the present application.

In one exemplary embodiment, as shown in FIG. 2, the apparatus for risk assessment of an atmospheric tank area based on FTA-ANN comprises:

the model building module 1 is used for building a fault tree model, identifying dangerous sources in the area of the normal pressure storage tank and building to obtain the fault tree model;

the model mapping and training module 2 is used for establishing an FTA-ANN model, mapping the fault tree model into an ANN, and performing multiple training through the training model to construct an FTA-ANN model;

and the model test and evaluation module 3 is used for performance test and risk evaluation, predicting based on the FTA-ANN model to obtain a predicted result, and comparing the predicted result with the FTA predicted result to realize risk evaluation on the normal pressure storage tank.

Specific limitations regarding an atmospheric tank risk assessment device based on FTA-ANN can be found in the above description of the method for risk assessment of an atmospheric tank region based on FTA-ANN, and will not be described in detail herein. The modules in the atmospheric storage tank risk evaluation device based on the FTA-ANN can be all or partially realized by software, hardware and combinations thereof. The above modules may be embedded in hardware or independent of a processor in the electronic device, or may be stored in software in a memory in the electronic device, so that the processor may call and execute operations corresponding to the above modules.

Fig. 3 is a schematic structural view of an electronic device according to an exemplary embodiment of the present application.

In an exemplary embodiment, the method for risk assessment of an atmospheric storage tank region based on FTA-ANN described above may be applied to an electronic device 400 shown in fig. 3, where the electronic device 400 includes at least a processor 410, a memory 420, a communication bus 430, and a communication interface 440, and the structure of the electronic device is shown in fig. 3.

The processor 410 may be a general-purpose central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), a microprocessor, or may be one or more integrated circuits for implementing the aspects of the present Application, such as an Application-specific integrated circuit (ASIC), a programmable logic device (Programmable Logic Device, PLD), or a combination thereof. The PLD may be a complex programmable logic device (Complex Programmable Logic Device, CPLD), a Field programmable gate array (Field-Programmable Gate Array, FPGA), general array logic (Generic Array Logic, GAL), or any combination thereof.

Alternatively, the processor 410 may include one or more CPUs. The electronic device 400 may include a plurality of processors 410. Each of these processors 410 may be a single-Core Processor (CPU) or a multi-core processor (multi-CPU).

It is noted that the processor 410 may refer to one or more devices, circuits, and/or processing cores for processing data (e.g., computer program instructions).

Memory 420 may be, but is not limited to, read-Only Memory (ROM) or other type of static storage device that can store static information and instructions, random access Memory (Random Access Memory, RAM) or other type of dynamic storage device that can store information and instructions, electrically erasable programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), compact disc Read-Only Memory (Compact Disc Read-Only Memory) or other optical disk storage, optical disk storage (including compact disc, laser disc, optical disc, digital versatile disc, blu-ray disc, etc.), magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.

Alternatively, the memory 420 may be stand alone and coupled to the processor 410 via the communication bus 430; memory 420 may also be integrated with processor 410.

Communication bus 430 is used to transfer information between components (e.g., between the processor and the memory) and communication bus 420 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one communication bus is shown in figure B, but not only one bus or one type of bus.

The communication interface 440 is used for the electronic device 400 to communicate with other devices or communication networks. Communication interface 440 includes a wired communication interface or a wireless communication interface. The wired communication interface may be, for example, an ethernet interface. The ethernet interface may be an optical interface, an electrical interface, or a combination thereof. The wireless communication interface may be a wireless local area network (Wireless Local Area Networks, WLAN) interface, a cellular network communication interface, a combination thereof, or the like.

In some embodiments, the electronic device 400 may also include an output device 450 and an input device 460 (not shown in fig. 1). The output device 450 communicates with the processor 410 and may display information in a variety of ways. For example, the output device 450 may be a liquid crystal display (Liquid Crystal Display, LCD), a light emitting diode (Light Emitting Diode, LED) display device, a Cathode Ray Tube (CRT) display device, or a projector (projector), or the like. The input device 460 is in communication with the processor 410 and may receive user input in a variety of ways. For example, the input device 460 may be a mouse, a keyboard, a touch screen device, a sensing device, or the like.

In some embodiments, memory 420 is used to store a computer program that performs aspects of the present application, and processor 410 may execute the computer program stored in memory 420. For example, the electronic device 400 may invoke and execute a computer program stored in the memory 420 via the processor 410 to implement the steps of the FTA-ANN based atmospheric tank zone risk assessment method provided by embodiments of the present application.

It should be understood that the method for evaluating risk of the atmospheric storage tank based on the table FTA-ANN provided by the application can be applied to an atmospheric storage tank risk evaluating device based on the table FTA-ANN, and the atmospheric storage tank risk evaluating device based on the table FTA-ANN can be implemented as part or all of the processor 410 in a mode of software, hardware or a combination of software and hardware, and is integrated in the electronic device 400.

In order to verify the advantages of the atmospheric storage tank risk evaluation method, the atmospheric storage tank risk evaluation device and the atmospheric storage tank risk evaluation electronic device based on the FTA-ANN provided by the application, as shown in fig. 9, the ANN model prediction result in some embodiments of the application is generally higher than the FTA prediction result for the fault tree analysis result, the correlation between input variables, namely the correlation between basic events, is considered by the ANN, a bootstrap method is used in the aspect of data acquisition, the bootstrap method can expand the sample size from an original sample according to a sampling-back method, and the small-scale subsample test evaluation problem is well solved. Meanwhile, the dynamic fault tree and the static fault tree are integrated into the ANN network, so that the defect that the static fault tree cannot describe the dynamic behavior of the system failure is overcome. The research model has a simple structure, and is visual and clear in input and output, and the input layer and the output layer of the ANN network can be quickly built as long as the fault tree model is built. The accuracy and stability of the method are better than those of the traditional deep learning method and BN network, and the risk assessment result is more scientific. From the sixth graph, several factors with the greatest influence on the leakage accident of the storage tank are defects of the external environment reaching the corrosion standard, deformation and damage of the storage tank, low enterprise management efficiency, staff training and consciousness, and conform to the evaluation result in the safety evaluation report. These aspects are therefore important to note for tank leak events.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The above examples represent only a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.

Claims

1. The atmospheric storage tank regional risk evaluation method based on FTA-ANN is characterized by comprising the following steps of:

building a fault tree model, identifying dangerous sources in the area of the normal pressure storage tank, and building to obtain the fault tree model;

and performing performance test and risk assessment, predicting based on the FTA-ANN model to obtain a prediction result, and comparing the prediction result with the FTA prediction result to realize risk assessment on the normal pressure storage tank area.

2. The FTA-ANN based atmospheric tank zone risk assessment method of claim 1 wherein a majority of the causes of a tank fire or explosion are due to tank leaks, and wherein a tank leak event is identified in the fault tree model as a roof event.

3. The FTA-ANN based atmospheric tank zone risk assessment method of claim 2, wherein the mapping rules mapping the fault tree model into an ANN comprise:

4. The FTA-ANN based atmospheric tank zone risk assessment method of claim 1, further comprising, prior to the training a plurality of times, effectively utilizing the failure samples and the number of samples in the failure tree model. When the availability of the sample data is insufficient, the sample needs to be collected again; when the number of samples is insufficient, the available data is randomly generated by using a bootstrap method, and then calculation is performed to obtain preliminary data.

5. The FTA-ANN based atmospheric storage tank zone risk assessment method of claim 1, wherein the training model comprises one input layer, two hidden layers and one output layer.

6. The method for evaluating risk of an atmospheric storage tank area based on FTA-ANN according to claim 5, wherein the number of nodes of two hidden layers is selected according to the method of hiding the number of neurons in combination with the characteristics of the fault tree model, and the selection rule comprises:

7. The method for evaluating risk of an atmospheric storage tank region based on FTA-ANN of claim 6, wherein the training model is trained by a back propagation algorithm, an optimizer thereof adopts Adam algorithm, uses mean square error as a loss function, and uses a method of small batch random gradient descent for optimization during training, the data is divided into small batches for training, and the maximum iteration number and the MSE threshold during the iteration are set to monitor the model performance. When the MSE has not improved for 10 consecutive iterations, the training process will stop.

8. The method for evaluating risk of an atmospheric storage tank area based on FTA-ANN according to claim 1, wherein in the performance test and risk evaluation, the application judges the influence of each accident factor on the failure of a top-level event by performing sensitivity analysis on input variables, and the calculation formula is as follows:

9. An atmospheric storage tank regional risk evaluation device based on FTA-ANN, which is characterized by comprising:

10. An electronic device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when calling and executing the computer program from the memory, implements the steps of the method according to any of the preceding claims 1-9.