Cross-scene safety emergency model construction method

Abstract

The invention provides a cross-scene safety emergency model construction method, and the method comprises the steps: building an observation point in an environment field, collecting the physical quantity fluctuation characteristics based on a time sequence under a disaster occurrence state through the observation point, obtaining a disaster feature code, and constructing a safety emergency model. According to the technical scheme, the composite upper concept of disasters and the characteristics of the disasters in physical representation are fully considered, and the safety emergency model is constructed based on the characteristics. And the possibility of fuzzy judgment is provided for disasters. Furthermore, the construction method constructs a model based on disaster characterization, and can be used for effectively updating an existing safety emergency rule set.

Description

Technical Field

[0001] The present invention relates to the field of information and communication technology specifically suitable for administrative, commercial, financial, management or supervisory purposes, and in particular to a method for constructing a cross-scenario security emergency model. Background Art

[0002] Currently, common safety emergency models rely on static rules or predefined decision trees for emergency response. However, static rules cannot cope with the complex changes between different disasters and the dynamic development of emergencies. The rule set can only cover known situations. For some emergencies or complex environments, traditional rule sets are prone to decision-making errors or delays.

[0003] To address these challenges, some emergency systems have begun to introduce reinforcement learning, allowing models to conduct trial and error learning in disaster simulations and gradually optimize decision-making strategies. Reinforcement learning can learn the optimal strategy through interaction with the environment. Some emergency systems have also introduced meta-learning for rapid adaptation. However, although meta-learning can increase the speed at which the model adapts to new disaster scenarios, its generalization ability depends on the similarity of the model between different disaster types. If the new disaster is significantly different from previous disasters in dynamics and structure, the meta-learning model may still not be able to adapt effectively. Meta-learning relies on part of historical data to complete adaptation to the new environment. When data is incomplete or of poor quality, its effectiveness is limited.

[0004] Disaster emergency scenarios require immediate decision-making, but reinforcement learning takes a long time to converge. The model needs to learn repeatedly in a simulated environment to find the best strategy and cannot quickly adapt to new environments. When facing new emergency disasters or unprecedented scenarios, it is often more important to make a general judgment on the scene rather than a relatively accurate classification judgment. Although existing meta-learning improves the adaptability of the model in new scenarios, it is difficult to make a general judgment on the higher-level concept of disaster, which is composed of different physical representations.

[0005] Therefore, it is necessary to provide a cross-scenario safety emergency model construction method to solve the above technical problems. Summary of the invention

[0006] The present invention aims to solve the problem that disaster emergency scenarios in the prior art require instant decision-making, but the convergence time of reinforcement learning is long, and the model needs to learn repeatedly in a simulated environment to find the best strategy, and cannot quickly adapt to the new environment. When facing new emergency disasters or unprecedented scenarios, it is often more important to make a general judgment on the scene and a relatively accurate classification judgment. Although the existing meta-learning improves the adaptability of the model in new scenarios, it is difficult to make a general judgment on the higher-level concept of disaster composed of different physical representations. A cross-scenario safety emergency model construction method is provided, which solves the above problem by judging the larger concept of disaster based on the characteristics of the suddenness of the disaster.

[0007] The present invention provides a cross-scenario safety emergency model construction method, comprising the following steps: S1. Collect basic physical data of different usage scenarios and construct several basic environment fields of detection scenarios for each usage scenario; S2. Establish basic physical scenarios based on basic physical data in each basic environmental field; S3, recording the data status under the basic environment field based on time series according to different physical observation attributes; S4, obtaining continuous stable data produced by the physical observation attribute based on the time series as the benchmark data; S5. Construct input points in the basic environment field according to the installation positions of the collection devices, and input the collected data; S6. Obtain instantaneous fluctuations of physical observation attributes based on collected data, and correspond the fluctuation characteristics of the physical observation attributes based on the benchmark data to the physical situation of the collected data; S7, integrating the physical observation properties of the basic environmental fields of each scene under the same acquisition scene, and obtaining the corresponding comprehensive physical reflection under the actual situation of the acquisition scene; S8. Normalize the corresponding comprehensive physical response in real scenarios according to the types of safety emergency problems.

[0008] According to the characteristics of disasters, disasters are often sudden, instantaneous, and highly severe, so the judgment of the broad concept of disasters should not be based on specific physical data, but on the characteristics of data fluctuations. The disaster type can then be inferred based on the characteristics of data fluctuations.

[0009] The cross-scenario safety emergency model construction method described in the present invention is, as a preferred method, specifically in step S3: Several observation points are evenly set up in each basic environmental field, and the observation point data under each general state are collected in real time.

[0010] The cross-scenario safety emergency model construction method described in the present invention is, as a preferred method, the specific method of step S4 is: The observation point denoises the collected values ​​of the physical observation attributes under the general state, and uses the continuous data with the largest proportion after denoising and exceeding the first threshold as the benchmark data.

[0011] The cross-scenario safety emergency model construction method described in the present invention, as a preferred embodiment, step S5 includes S51, introducing data collected by the front-end system into the corresponding actual collection position in the basic environment field; S52. Each observation point collects the physical data fluctuations relative to the benchmark data caused by the front-end system data in real time.

[0012] The method for constructing a cross-scenario safety emergency model described in the present invention is, as a preferred embodiment, that the instantaneous fluctuation in step S6 is an instantaneous high-quantity difference disturbance.

[0013] The cross-scenario safety emergency model construction method described in the present invention is, as a preferred embodiment, the observation point observation data includes electrical data, optical data, mechanical data, thermal data, fluid data, chemical data, acoustic data and geophysical data.

[0014] The method for constructing a cross-scenario safety emergency model described in the present invention is, as a preferred method, the method for obtaining the type of safety emergency problem in step S8 is that the input signal is manually confirmed to obtain the safety emergency problem corresponding to the input signal, and the input data corresponding to the safety emergency problem is categorized and marked. The observation point converts the above signal into a data fluctuation feature and inherits the mark while corresponding to the timestamp to obtain the physical observation attribute data fluctuation feature when the safety emergency problem occurs in chronological order.

[0015] The cross-scenario safety emergency model construction method described in the present invention is, as a preferred method, observation points are distributed throughout the basic environment field based on a three-dimensional array.

[0016] The beneficial effects of the present invention are as follows: This technical solution fully considers the characteristics of the complex superordinate concept of disaster in physical representation, and builds a safety emergency model based on the above characteristics. It provides the possibility of fuzzy judgment for disasters. Furthermore, this construction method builds a model based on disaster representation, which can be used to effectively update the existing safety emergency rule set. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flow chart of a method for building a cross-scenario safety emergency response model. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Example 1

[0019] like Figure 1 As shown, a method for constructing a cross-scenario safety emergency model includes the following steps: S1. Collect basic physical data of different usage scenarios and construct several basic environment fields of detection scenarios for each usage scenario; S2. Establish basic physical scenarios based on basic physical data in each basic environmental field; S3, recording the data status under the basic environment field based on time series according to different physical observation attributes; S4, obtaining continuous stable data produced by the physical observation attribute based on the time series as the benchmark data; S5. Construct input points in the basic environment field according to the installation positions of the collection devices, and input the collected data; S6. Obtain instantaneous fluctuations of physical observation attributes based on collected data, and correspond the fluctuation characteristics of the physical observation attributes based on the benchmark data to the physical situation of the collected data; S7, integrating the physical observation properties of the basic environmental fields of each scene under the same acquisition scene, and obtaining the corresponding comprehensive physical reflection under the actual situation of the acquisition scene; S8. Normalize the corresponding comprehensive physical response in real scenarios according to the types of safety emergency problems.

[0020] According to the above construction method, when a disaster occurs, this model can effectively reflect the disaster and judge whether the disaster is a preset disaster by whether the current physical characteristics have been predefined based on the rule set. According to this method, the rule set can be effectively broadened, and the fuzzy concept of disaster can be judged based on the instantaneous fluctuation itself. That is, all physical fluctuations with instantaneous high quantitative differences may be defined as disasters.

[0021] The specific method of step S3 is: Several observation points are evenly set up in each basic environmental field, and the observation point data under each general state are collected in real time.

[0022] The specific method of step S4 is: The observation point denoises the collected values ​​of the physical observation attributes under the general state, and uses the continuous data with the largest proportion after denoising and exceeding the first threshold as the benchmark data.

[0023] Step S5 includes S51, introducing data collected by the front-end system into the corresponding actual collection position in the basic environment field; S52. Each observation point collects the physical data fluctuations relative to the benchmark data caused by the front-end system data in real time.

[0024] The instantaneous fluctuation in step S6 is an instantaneous high quantity difference disturbance.

[0025] The observation data at the observation points include electrical data, optical data, mechanical data, thermal data, fluid data, chemical data, acoustic data and geophysical data.

[0026] The method for obtaining the type of safety emergency problem in step S8 is that the input signal is manually confirmed to obtain the safety emergency problem corresponding to the input signal, and the input data corresponding to the safety emergency problem is categorized. The observation point converts the above signal into data fluctuation characteristics and inherits the mark while corresponding to the timestamp to obtain the physical observation attribute data fluctuation characteristics when the safety emergency problem occurs in chronological order.

[0027] The observation points are distributed throughout the basic environment field based on a three-dimensional array.

[0028] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for constructing a cross-scenario safety emergency model, characterized by: The following steps are involved: S1. Collect basic physical data of different usage scenarios and construct several basic environment fields of detection scenarios for each usage scenario; S2. establishing a basic physical scenario in each basic environment field according to the basic physical data; S3, recording the data state under the basic environment field based on time series according to different physical observation attributes; S4, obtaining continuous stable data produced by the physical observation attribute based on the time series as the benchmark data; S5. Construct input points in the basic environment field according to the installation positions of the collection devices, and input the collected data; S6. Obtain instantaneous fluctuations of physical observation attributes according to the collected data, and correspond the fluctuation characteristics of the physical observation attributes to the reference data with the physical conditions of the collected data; S7, integrating the physical observation properties of the basic environment fields of each scene under the same acquisition scene to obtain the corresponding comprehensive physical reflection under the actual situation of the acquisition scene; S8. Normalize the corresponding comprehensive physical response in real scenarios according to the type of safety emergency problem.

2. A cross-scenario safety emergency model construction method according to claim 1, characterized in that: The specific method of step S3 is: Several observation points are evenly arranged in each of the basic environmental fields, and the observation point data under each general state are collected in real time.

3. A cross-scenario safety emergency model construction method according to claim 2, characterized in that: The specific method of step S4 is: The observation point denoises the collected values ​​of the physical observation attribute under a general state, and uses the continuous data with the largest proportion after denoising and exceeding a first threshold as the benchmark data.

4. A cross-scenario safety emergency model construction method according to claim 1, characterized in that: The step S5 comprises: S51, introducing data collected by the front-end system into the corresponding actual collection position in the basic environment field; S52. Each observation point collects the physical data fluctuations relative to the benchmark data caused by the front-end system data in real time.

5. According to claim 1, a cross-scenario safety emergency model construction method is characterized by: The instantaneous fluctuation in step S6 is an instantaneous high quantity difference disturbance.

6. A cross-scenario safety emergency model construction method according to claim 2, characterized in that: The observation data at the observation points include electrical data, optical data, mechanical data, thermal data, fluid data, chemical data, acoustic data and geophysical data.

7. A cross-scenario safety emergency model construction method according to claim 1, characterized in that: The method for obtaining the type of safety emergency problem in step S8 is that the input signal is manually confirmed to obtain the safety emergency problem corresponding to the input signal, and the input data corresponding to the safety emergency problem is categorized. The observation point converts the above signal into a data fluctuation feature and inherits the mark while corresponding to the timestamp to obtain the physical observation attribute data fluctuation feature when the safety emergency problem occurs in chronological order.

8. A cross-scenario safety emergency model construction method according to claim 2, characterized in that: The observation points are distributed throughout the basic environment field based on a three-dimensional array.

Images