Gas station monitoring and early warning system based on artificial intelligence

Through a multi-module collaborative working system based on artificial intelligence, a scenario division and discrimination model is built, and the problem of single early warning mode in the gas station monitoring and early warning system is solved, and refined early warnings are achieved for different risk levels and scenarios are improved, and the accuracy of early warning is improved. Sex and real-time.

CN118570983BActive Publication Date: 2024-11-08SHANDONG YUELAI DIGITAL TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410541779.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-04-30
Publication Date
2024-11-08
Estimated Expiration
2044-04-30

AI Technical Summary

Technical Problem

The existing gas station monitoring and early warning system has a single warning mode and cannot provide refined early warnings for different hazard levels and scenarios, resulting in a reduction in early warning effect and real-time performance.

Method used

Adopt a multi-module collaborative working system based on artificial intelligence, obtain real-time data by collecting modules, storing modules data, processing modules build scenario division and discrimination models, integrate modules comprehensive early warning mode and dangerous data, and the early warning module sends corresponding alarm signals to achieve the correct Refined early warnings for different risk levels and scenarios.

Benefits of technology

It improves the accuracy and real-time nature of early warning, realizes all-round and real-time monitoring and early warning of gas stations, enhances the adaptability and flexibility of the early warning system, and ensures the refined early warning effect for different hazard levels and scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118570983B_ABST
    Figure CN118570983B_ABST
Patent Text Reader

Abstract

The present invention discloses a gas station monitoring and early warning system based on artificial intelligence, and relates to the field of monitoring and early warning technology. It includes a collection module: used to obtain real-time detection data; a storage module: used to store detection data, and the detection data includes equipment data, environmental data and behavior data. The present invention improves the accuracy and real-time performance of early warning by adopting multi-module collaborative work, realizes all-round and real-time monitoring and early warning of gas stations, builds a processing model through detection data, improves the comprehensiveness of the early warning system, completes the scene division of gas stations, and outputs early warning modes and dangerous data according to the scene and real-time detection data. The setting improves the early warning effect and real-time performance of the early warning system to a certain extent, and solves the problem that the early warning mode is single and cannot be refined for different danger levels and scenes during the early warning process.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of monitoring and early warning technology, and in particular to a gas station monitoring and early warning system based on artificial intelligence. Background Art

[0002] With the development of science and technology and the progress of society, gas stations, as an important part of urban infrastructure, are receiving more and more attention for their operational safety and management efficiency. However, in the actual operation of gas stations, since gasoline in gas stations is flammable and explosive, in order to ensure safety, it is necessary to design a gas station monitoring and early warning system based on artificial intelligence to monitor and warn the internal equipment, operators, vehicles and personnel of the gas station.

[0003] After searching, the Chinese invention patent with publication number "CN112258042A" discloses "an artificial intelligence-based gas station key area and personnel safety hazard monitoring and early warning system and method". The application uses a video surveillance module to monitor video information such as personnel and vehicle license plates, and a sensor group to monitor equipment operation status and temperature and other parameter information, and uses a high-speed computing communication component to preliminarily process the collected video information and parameter information and transmit it to the AI ​​computing processing platform; the AI ​​processing platform first stores and saves the collected video information and parameter information, and then compares the video content and parameter status with the preset reference, as well as the setting of graded alarms. It can monitor and alarm the monitoring equipment and personnel activities in real time, identify safety hazards, and warn and report possible safety hazards.

[0004] In addition, the Chinese invention patent with publication number "CN117610821A" discloses "an artificial intelligence-based control and warning system and method". The application screens the target vehicle and the corresponding target transition area based on the collected data, selects the best allocation plan, and controls the target vehicle. When an abnormal situation occurs, it will issue an early warning to relevant technical personnel to save costs.

[0005] Finally, the Chinese invention patent with publication number "CN115249115A" discloses "an early warning system based on artificial intelligence". This application avoids gas leakage problems by marking multiple gas leakage sensors with serial numbers and placing them at different connections of the pipeline for gas monitoring. At the same time, the specific location of each sensor is registered and stored. When a gas leak occurs, it can provide timely warnings and help maintenance personnel to query the specific location in time, greatly reducing safety hazards.

[0006] However, when the above-mentioned early warning system and method and similar systems and methods are actually used, the early warning system has a single early warning mode and cannot provide refined early warning for different danger levels and scenarios, which reduces the early warning effect and real-time performance to a certain extent. Summary of the invention

[0007] The purpose of the present invention is to provide a gas station monitoring and early warning system based on artificial intelligence to solve the problems raised in the above background technology.

[0008] To achieve the above purpose, the present invention provides the following technical solution: a gas station monitoring and early warning system based on artificial intelligence, comprising:

[0009] Collection module: used to obtain real-time detection data;

[0010] Storage module: used to store detection data, which includes equipment data, environment data and behavior data;

[0011] Processing module: used to read any two items of detection data and build an artificial intelligence-based processing model. The processing model divides gas stations into different scenarios to obtain different early warning modes, and determines the danger level of real-time detection data and outputs danger data;

[0012] Integration module: used to integrate early warning patterns and hazard data, and output feedback results;

[0013] The early warning module is used to issue different alarm commands according to the feedback results;

[0014] The processing model includes:

[0015] Based on the characteristics of the detection data, the scene data of the gas station is obtained, a number of influence data are derived based on the scene data, a discrimination method is established, the scene data and the influence data are associated, the influence of the influence data on the scene data is discriminated, and different early warning modes are output based on the degree of influence;

[0016] Read the real-time detection data, build a rating method, select scenario data with the same characteristics as the real-time detection data, compare the similarity between the real-time detection data and the impact data, and output different danger data based on the degree of similarity and the scenario data.

[0017] As a further preferred embodiment of the present technical solution, the discrimination method includes: limiting the unification of scene data and influencing data features, and generating real-time scene modeling based on the scene data;

[0018] Read the impact data and generate it in the real-time scene modeling, collect various attribute data of the real-time scene modeling when the impact data does not appear, and record it as data A, retrieve various attribute data of the real-time scene modeling when the impact data is generated, and record it as data B;

[0019] The midpoint value of each attribute data is set as the threshold, and the data difference between data B and data A is compared. Based on the size of the data difference and the threshold comparison, different warning modes are output.

[0020] As a further preferred embodiment of the present technical solution, the rating method includes: extracting features within the real-time detection data, retrieving scene data with the same features, continuously inputting multiple segments of continuous real-time detection data and performing feature vector comparison with all influencing data, selecting influencing data that is close to the feature vectors of the multiple segments of continuous real-time detection data, and outputting them as associated data, and the associated data outputs different danger data based on the scene data.

[0021] As a further preferred embodiment of the present technical solution, when the processing model is constructed based on the device data and the environment data, the processing model includes:

[0022] Based on the characteristics of equipment data and environmental data, the layout positions of internal equipment in the gas station and the layout positions of the gas station are obtained. Based on the layout positions of the gas station, several influencing positions are derived. Based on the discrimination method, different early warning modes are generated based on the distance difference between the influencing positions and the layout positions of the gas station.

[0023] By reading the equipment data in the real-time detection data, the danger level of the layout position of the internal equipment of the gas station in several influencing positions in the gas station is evaluated, and different danger data are output.

[0024] As a further preferred embodiment of the present technical solution, when the processing model is constructed based on the device data and the behavior data, the processing model includes:

[0025] Based on the equipment data and behavior data, the operating status of the equipment in the gas station and the behavior of the personnel and vehicles in the gas station are obtained. Based on the operating status of the equipment in the gas station, several influencing actions are derived. Based on the discrimination method, different early warning modes are constructed based on the differences between the influencing actions and the behavior actions.

[0026] By reading the behavioral data in the real-time detection data, the danger level of the behavioral actions of people and vehicles in the gas station in several influencing actions is evaluated, and different danger data are output.

[0027] As a further preferred embodiment of the present technical solution, when the processing model is constructed based on the environmental data and the behavioral data, the processing model includes:

[0028] Based on environmental data and behavioral data, we obtain environmental parameters in gas stations and the behaviors of people in gas stations. Based on environmental parameters, we derive several influencing factors. Based on the discrimination method, we build different early warning modes based on the differences between influencing factors and environmental data.

[0029] By reading real-time detection data, the danger level of environmental factors in the gas station among several influencing factors is evaluated and different danger data are output.

[0030] As a further preferred embodiment of the present technical solution, the integration module includes: establishing a high-weight integration correction rule that is biased towards the early warning mode, and the integration correction rule corrects the output feedback result based on the early warning mode and the danger level.

[0031] As a further preferred embodiment of the present technical solution, the early warning module classifies different early warning modes according to the feedback results and sends out corresponding alarm signals, and the alarm signals include any one or a combination of sound signals, light signals, vibration signals and image signals.

[0032] As a further preferred embodiment of the present technical solution, the collection module collects real-time equipment data for understanding the layout and operating status of the internal equipment of the gas station, the collection module collects real-time environmental data for monitoring the environmental conditions within the gas station, and the collection module collects real-time behavior data for observing the behavior of people and vehicles within the gas station.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The gas station monitoring and early warning system based on artificial intelligence improves the accuracy and real-time performance of early warning by adopting multi-module collaborative work, realizes all-round and real-time monitoring and early warning of gas stations, builds processing models through detection data, improves the comprehensiveness of the early warning system, completes the scene division of gas stations, and outputs early warning modes and dangerous data according to scenes and real-time detection data, so that the early warning effect and real-time performance of the early warning system are improved to a certain extent, solving the problem that the early warning mode is single and cannot be refined for different danger levels and scenes during the early warning process.

[0035] In addition, the system obtains the scene data of the gas station by processing the model based on the characteristics of the detection data, and derives the influence data according to the scene data. By establishing the setting of the discrimination method, the scene data and the influence data are associated, the influence of the influence data on the scene data is discriminated, and the three warning modes of high, medium and low are output based on the degree of influence, so that the warning system has higher adaptability and real-time performance. By outputting three types of danger data, strong, medium and weak, according to the similarity between the real-time detection data and the influence data, the warning ability of the warning system is further improved.

[0036] Furthermore, during actual operation, the monitoring and early warning system proposed in the present application establishes high-weight integrated correction rules that are biased towards the early warning mode through the integration module, and corrects the output feedback results, so that the early warning system has higher accuracy and flexibility. By setting up the early warning module, different early warning modes can be classified during actual operation, and corresponding alarm signals can be issued, thereby improving the warning effect of the early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is the operation logic diagram of the system of the present invention;

[0038] Figure 2 is an operation diagram of the association rules of the present invention;

[0039] Figure 3 It is an operation diagram of the discrimination method of the present invention;

[0040] Figure 4 is an operation diagram of the rating method of the present invention;

[0041] Figure 5 An operational diagram for integrating corrective rules for the present invention;

[0042] Figure 6 It is an operation diagram of the first embodiment of the processing model of the present invention;

[0043] Figure 7 It is an operation diagram of the second embodiment of the processing model of the present invention;

[0044] Figure 8 This is an operation diagram of the third embodiment of the processing model of the present invention. DETAILED DESCRIPTION

[0045] 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 only 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.

[0046] The gas station monitoring and early warning system provided by this application not only relies on advanced hardware equipment and software systems in actual operation, but also requires efficient personnel cooperation and management strategies. It should be added that several aspects need to be paid attention to when the system proposed by this application is running:

[0047] First, hardware equipment maintenance: ensure the normal operation of various monitoring equipment, such as cameras, hard disk recorders, alarms, etc., regularly check the equipment connection lines and power supplies to prevent equipment failures caused by line aging or unstable power supply, regularly clean and maintain the equipment to avoid dust or dirt accumulation that affects equipment performance, and regularly replace vulnerable parts such as camera lenses, alarm sensors, etc. to ensure the stable operation of the monitoring system;

[0048] Second, software system optimization: The monitoring and early warning system software needs to be customized and optimized according to actual needs to ensure its efficiency and compatibility, regularly update the software version, fix vulnerabilities, and improve system security and stability;

[0049] Third, personnel management: Provide professional training for monitoring and early warning system operators to ensure that they are proficient in the use of the system and precautions, formulate clear job responsibilities and operating procedures, standardize the operators' operating behaviors, prevent misoperation or abuse of authority, establish a complete duty system, ensure that the monitoring and early warning system is manned around the clock, and handle emergencies in a timely manner.

[0050] like Figure 1 As shown, the present invention provides a technical solution: a gas station monitoring and early warning system based on artificial intelligence, including: a collection module: used to obtain real-time detection data, a storage module: used to store detection data, the detection data includes equipment data, environmental data and behavior data. It should be noted that in the present application, the collection module collects real-time equipment data for understanding the layout and operating status of the internal equipment of the gas station, the collection module collects real-time environmental data for monitoring the environmental conditions in the gas station, the collection module collects real-time behavior data for observing the behavior of people and vehicles in the gas station, and the processing module is used to read any two data in the detection data and construct a processing model based on artificial intelligence. The processing model divides the gas station according to the scene to obtain different early warning modes and performs danger level classification on the real-time detection data. Identify and output the dangerous data, integration module: used to integrate the early warning mode and dangerous data, and output the feedback results, early warning module, used to issue different alarm commands according to the feedback results. It should be noted that in the present application, the processing model includes: obtaining the scene data of the gas station based on the characteristics of the detection data, deriving a number of influence data based on the scene data, establishing a discrimination method, associating the scene data and the influence data, discriminating the influence of the influence data on the scene data, and outputting three early warning modes of high, medium and low based on the degree of influence, reading real-time detection data, constructing a rating method, selecting scene data with the same characteristics as the real-time detection data, comparing the similarity between the real-time detection data and the influence data, and outputting three types of dangerous data of strong, medium and weak based on the degree of similarity and the scene data.

[0051] It should be added that in the present application, when the processing model derives several influencing data based on the scenario data, first, the processing model cleans and pre-processes the collected scenario data to ensure the quality and consistency of the data. Next, the scenario data is subjected to feature engineering processing to extract the features that have an impact on the safety status of the gas station. Then, the association rules between the scenario data and the potential influencing factors are established, and the laws and patterns therein are discovered using data mining and pattern recognition techniques. On this basis, the potential influencing factors are defined and classified to form a set of influencing data. Next, the scenario data is further analyzed and processed according to the defined set of influencing data to derive data related to each influencing factor. The derived influencing data needs to be associated and evaluated with the scenario data to determine their actual impact on the safety status of the gas station. Finally, the processing model outputs the derived influencing data.

[0052] It should be further added that reference Figure 2 It can be seen that in this application, the association rules include: discovering possible potential association relationships from scene data through exploratory analysis and data mining technology. It should be noted that these association relationships describe the correlation between different variables, that is, when certain variables appear at the same time, other variables may also appear at the same time. For example, in the gas station monitoring and early warning system, by analyzing scene data such as equipment status, environmental parameters and personnel activities, potential influencing factors related to the safety status can be discovered, such as equipment failure, environmental anomalies, illegal behavior, etc. By establishing these association relationships, it can help understand and explain the complex relationship between scene data and guide subsequent data processing and analysis work. In addition, it should be added that based on these association relationships, the scene data can be converted into influencing data, and then the factors that have potential impact on the safety status of the gas station can be identified.

[0053] refer to Figure 3 It can be seen that in the present application, the discrimination method includes: limiting the unity of scene data and influencing data features, generating real-time scene modeling based on scene data; reading influencing data and generating it in real-time scene modeling, collecting various attribute data of real-time scene modeling when no influencing data appears, and recording it as data A, retrieving various attribute data of real-time scene modeling when influencing data is generated, and recording it as data B; setting the midpoint value of each attribute data as the threshold, comparing the data difference between data B and data A, and based on the size of the comparison between the data difference and the threshold, outputting three warning modes of high, medium and low.

[0054] It should be noted that in the present application, the warning modes are: high warning mode, medium warning mode and low warning mode. The high warning mode indicates that the safety status of the gas station is in extreme danger and immediate measures need to be taken to deal with it. The medium warning mode indicates that the safety status of the gas station is in general danger and needs to be closely monitored and timely measures need to be taken. The low warning mode indicates that the safety status of the gas station is relatively stable, but vigilance is still required.

[0055] It should be further added that in actual applications, the early warning mode in this application will issue corresponding alarm commands for different early warning modes based on the feedback results output by the integration module. For example, in the high early warning mode, the early warning module will immediately issue an emergency alarm to notify relevant personnel to take emergency measures. In the medium early warning mode, the early warning module will issue a warning signal to remind relevant personnel to pay attention to the safety status of the gas station. In the low early warning mode, the early warning module will issue a normal early warning signal to prompt relevant personnel to remain vigilant.

[0056] It should be noted that the reference Figure 6 It can be seen that in this embodiment, when the processing model is constructed based on equipment data and behavior data, the processing model will obtain the operating status of the equipment in the gas station and the behavioral actions of the personnel and vehicles in the gas station based on the equipment data and behavior data, and derive a number of influencing actions based on the operating status of the equipment in the gas station, and then based on the discrimination method, construct three early warning modes of high, medium and low formed by the differences between the influencing actions and the behavioral actions, and evaluate the danger level of the behavioral actions of the personnel and vehicles in the gas station in several influencing actions by reading the behavioral data in the real-time detection data, and output three types of danger data: strong, medium and weak. It should be noted that in this embodiment, the collection module for obtaining real-time detection data is specifically attached. Figure 6 The detector appears in the monitor, which is responsible for capturing the information of the surrounding environment in real time and transmitting this information to the storage module in the form of digital signals. The storage module is a storage element set inside the monitor. Figure 6 The processing module in this embodiment is specifically a terminal server, which is located in a maintenance station in a gas station. Figure 6It can be seen that in this embodiment, since various equipment such as oil pumps and oil tanks are installed in the gas station, the operating status of oil pumps, oil tanks and other equipment is detected in real time through detectors such as temperature sensors, and the behavior of personnel and vehicles, such as walking, stopping, refueling, and whether there is static electricity discharge during refueling, are captured through detectors such as monitors and other sensors. Therefore, in actual operation, the processing model mainly determines the impact of the operating status of equipment in the gas station and the behavior of personnel and vehicles on the safety status of the gas station, and timely discovers potential safety hazards based on real-time monitoring data and early warning modes, such as equipment failure or abnormal behavior around the equipment, and takes corresponding measures, such as stopping equipment operation, alarming, and notifying management personnel, to ensure the safe operation of the gas station. It should be noted that the object protected by this embodiment is the equipment, and in this application, the early warning mode is specifically embodied in the specific equipment in the gas station, with the specific equipment as the main body, and the early warning mode is divided according to the proximity to the specific equipment.

[0057] In addition, reference Figure 6 and Figure 7 It can be seen that in this embodiment, when the processing model is constructed based on equipment data and environmental data, the processing model will obtain the layout position of the internal equipment of the gas station and the layout position of the gas station based on the characteristics of the equipment data and environmental data, and derive several influencing positions based on the layout position of the gas station, and then based on the discrimination method, construct three warning modes of high, medium and low generated by the distance difference between the influencing position and the layout position of the gas station, and evaluate the danger level of the layout position of the internal equipment of the gas station among several influencing positions in the gas station by reading the equipment data in the real-time detection data, and output three types of danger data of strong, medium and weak. It should be supplemented that, in this embodiment, the processing model obtains the layout position of the gas station by reading the equipment data of the gas station, and obtains the layout position of the gas station through the detector. Therefore, when the processing module of this embodiment is running, refer to Figure 7 It can be seen that when the position of the fire-fighting equipment changes, three warning levels will be formed outside the gas pump through the discrimination method, and the marked fire-fighting equipment will be rated for the dangerous data level based on the position of the fire-fighting equipment and the gas pump when the position of the fire-fighting equipment changes. It should be noted that in actual use, the fire-fighting equipment can be monitored and identified in real time through the detector. It should also be added that in this embodiment, when the processing model is actually running, it mainly judges the impact of the layout position of the equipment in the gas station on the safety of the gas station. According to the real-time monitoring data and warning mode, the potential safety hazards in the equipment location are discovered in time, and corresponding measures are taken, such as adjusting the equipment location, increasing safety protection measures, etc., to ensure the safe operation of the gas station.

[0058] refer to Figure 6 and Figure 8It can be seen that in this embodiment, when the processing model is constructed based on environmental data and behavioral data, the processing model will obtain the environmental parameters in the gas station and the behavioral actions of the personnel in the gas station based on the environmental data and behavioral data, and derive a number of influencing factors based on the environmental parameters, and then based on the discrimination method, construct three early warning modes of high, medium and low formed due to the differences between the influencing factors and the environmental data; by reading the real-time detection data, the danger level of the environmental factors in the gas station among the several influencing factors is evaluated, and three types of danger data, strong, medium and weak, are output. At this time, the processing model mainly judges the impact of the environmental factors and personnel behavior in the gas station on the safety status of the gas station during actual operation. It should be added that in this embodiment, the processing model installs various sensors (i.e., collection modules) in the gas station, such as gas detectors, temperature sensors, cameras, etc. , to monitor environmental parameters such as gas concentration, temperature, and personnel activities in real time, and capture personnel behaviors through a camera monitoring system, such as whether they are wearing a safety helmet or whether they are active in a prohibited area. Since the processing model continuously reads real-time detection data, including environmental parameters and personnel behaviors, during actual operation, the processing model mainly determines the impact of environmental factors and personnel behaviors in the gas station on the safety status of the gas station. By using real-time monitoring data and early warning modes, potential safety hazards are discovered in a timely manner, and corresponding measures are taken, such as reminding personnel to pay attention to safety, alarming and notifying management personnel, etc., to ensure the safe operation of the gas station. It should be added that the main monitored and protected objects in this embodiment are personnel, so the generated early warning mode covers the entire gas station, and the personnel in the gas station output different levels of danger data based on their location and their own behavior.

[0059] refer to Figure 4 It can be seen that in the present application, the rating method includes: extracting features within real-time detection data, retrieving scene data with the same features, continuously inputting multiple segments of continuous real-time detection data and performing feature vector comparison with all influencing data to select influencing data that is close to the feature vectors of multiple segments of continuous real-time detection data, and outputting them as associated data. The associated data outputs three types of danger data: strong, medium, and weak, based on the scene data.

[0060] It should be noted that in this application, the risk data of the rating method are: strong risk data, medium risk data and weak risk data. Among them, strong risk data indicates that the safety status of the gas station is facing serious threats, which requires high attention and immediate measures. Medium risk data indicates that there is a certain risk in the safety status of the gas station, which requires close attention and timely measures; weak risk data indicates that the safety status of the gas station is relatively safe, but still requires continuous attention. In actual applications, the risk data output by the rating method is associated with the early warning mode, and then the feedback result is output through the integration module preset in this application.

[0061] It should be further explained that, in the present application, the integration module includes: establishing a high-weight integration correction rule that is biased towards the early warning mode, the integration correction rule takes the early warning mode as the main body and the danger level as the reference to correct the output feedback result;

[0062] refer to Figure 5 It can be seen that in this application, the integrated correction rules include: when the warning mode is high, if the danger level is weak, the warning mode is adjusted to medium; when the warning mode is medium, if the danger level is strong, the warning mode is adjusted to high; if the danger level is weak, the warning mode is adjusted to low; when the warning mode is low, if the danger level is strong or medium, the warning mode is adjusted to medium; if the warning mode and the danger level are consistent, the original warning mode remains unchanged.

[0063] It should also be added that during the actual operation of the present application, the early warning module classifies different early warning modes according to the feedback results and sends out corresponding alarm signals. The alarm signals include any one or a combination of sound signals, light signals, vibration signals and image signals.

[0064] The artificial intelligence-based gas station monitoring and early warning system proposed in this application, during actual operation, the collection module continuously obtains real-time equipment data, environmental data and behavioral data to ensure that the system has the latest monitoring information. These data are then saved by the storage module to provide support for subsequent processing. The processing module then starts to run, reads the detection data and builds a processing model based on artificial intelligence, analyzes and identifies the data, generates corresponding early warning patterns and danger data, and then the integration module integrates various early warning patterns and danger data, outputs the final feedback results, and optimizes and adjusts the results by integrating correction rules. After the feedback results are input into the early warning module, corresponding alarm signals are triggered, including sound, light, vibration and image signals, as well as possible combinations, to warn the monitoring personnel.

[0065] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. The gas station monitoring and early warning system based on artificial intelligence is characterized by: include: Collection module: used to obtain real-time detection data; Storage module: used to store detection data, which includes equipment data, environment data and behavior data; Processing module: used to read any two items of detection data and build an artificial intelligence-based processing model. The processing model divides gas stations into different scenarios to obtain different early warning modes, and determines the danger level of real-time detection data and outputs danger data; Integration module: used to integrate early warning patterns and hazard data, and output feedback results; The early warning module is used to issue different alarm commands according to the feedback results; The processing model includes: Based on the characteristics of the detection data, the scene data of the gas station is obtained, a number of influence data are derived based on the scene data, a discrimination method is established, the scene data and the influence data are associated, the influence of the influence data on the scene data is discriminated, and different early warning modes are output based on the degree of influence; Read the real-time detection data, build a rating method, select scenario data with the same characteristics as the real-time detection data, compare the similarity between the real-time detection data and the impact data, and output different danger data based on the degree of similarity and scenario data; The discrimination method includes: limiting the unification of scene data and influencing data features, and generating real-time scene modeling based on the scene data; Read the impact data and generate it in the real-time scene modeling, collect various attribute data of the real-time scene modeling when the impact data does not appear, and record it as data A, retrieve various attribute data of the real-time scene modeling when the impact data is generated, and record it as data B; Set the midpoint value of each attribute data as the threshold, compare the data difference between data B and data A, and output different warning modes based on the size of the data difference and the threshold comparison; The rating method includes: extracting features within real-time detection data, retrieving scene data with the same features, continuously inputting multiple segments of continuous real-time detection data and performing feature vector comparison with all influencing data, selecting influencing data that is close to the feature vectors of the multiple segments of continuous real-time detection data, and outputting the data as associated data, wherein the associated data outputs different danger data based on the scene data.

2. The artificial intelligence-based gas station monitoring and early warning system according to claim 1 is characterized by: When the processing model is constructed based on the device data and the environment data, the processing model includes: Based on the characteristics of equipment data and environmental data, the layout positions of internal equipment in the gas station and the layout positions of the gas station are obtained. Based on the layout positions of the gas station, several influencing positions are derived. Based on the discrimination method, different early warning modes are generated based on the distance difference between the influencing positions and the layout positions of the gas station. By reading the equipment data in the real-time detection data, the danger level of the layout position of the internal equipment of the gas station in several influencing positions in the gas station is evaluated, and different danger data are output.

3. The artificial intelligence-based gas station monitoring and early warning system according to claim 1 is characterized by: When the processing model is built based on the device data and the behavior data, the processing model includes: Based on the equipment data and behavior data, the operating status of the equipment in the gas station and the behavior of the personnel and vehicles in the gas station are obtained. Based on the operating status of the equipment in the gas station, several influencing actions are derived. Based on the discrimination method, different early warning modes are constructed based on the differences between the influencing actions and the behavior actions. By reading the behavioral data in the real-time detection data, the danger level of the behavioral actions of people and vehicles in the gas station in several influencing actions is evaluated, and different danger data are output.

4. The artificial intelligence-based gas station monitoring and early warning system according to claim 1 is characterized by: When the processing model is constructed based on the environmental data and the behavioral data, the processing model includes: Based on environmental data and behavioral data, we obtain environmental parameters in gas stations and the behaviors of people in gas stations. Based on environmental parameters, we derive several influencing factors. Based on the discrimination method, we build different early warning modes based on the differences between influencing factors and environmental data. By reading real-time detection data, the danger level of environmental factors in the gas station among several influencing factors is evaluated and different danger data are output.

5. The artificial intelligence-based gas station monitoring and early warning system according to claim 1 is characterized by: The integration module includes: establishing a high-weight integration correction rule that is biased towards the early warning mode, and the integration correction rule corrects the output feedback result based on the early warning mode and the danger level.

6. The artificial intelligence-based gas station monitoring and early warning system according to claim 1 is characterized by: The early warning module classifies different early warning modes according to the feedback results and sends out corresponding alarm signals, which include any one or a combination of sound signals, light signals, vibration signals and image signals.

7. The artificial intelligence-based gas station monitoring and early warning system according to claim 1 is characterized by: The collection module collects real-time equipment data for understanding the layout and operating status of the internal equipment of the gas station, the collection module collects real-time environmental data for monitoring the environmental conditions within the gas station, and the collection module collects real-time behavior data for observing the behavior of people and vehicles in the gas station.