Flame detector and its calibration and analysis system based on artificial intelligence
By introducing artificial intelligence technology into flame detectors, combining active self-test and real-time detection actions, using optical sensors and light sources for self-test and data optimization, the problem of flame detectors being susceptible to environmental interference is solved, and automatic Detection and parameter correction improve the accuracy and reliability of detection.
Patent Information
- Application Number
- CN202410513133.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-26
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-04-26
AI Technical Summary
Flame detectors are susceptible to temperature and humidity environments that lead to failures or false alarms, and on-site testing requires a lot of manpower and time resources, which poses a risk of missed inspection.
Design a flame detector based on artificial intelligence, including active self-test and real-time detection actions, self-test and real-time data analysis through optical sensors and light sources, and data optimization and parameter correction units are used to optimize data and parameter correction. Avoid external environment interference and achieve independent detection.
It realizes automatic detection and parameter correction without affecting normal operation, reduces false alarms and faults, avoids the time and labor consumption of manual detection, and improves the accuracy and reliability of detection.
Smart Images

Figure CN118424460B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of flame detectors, and in particular to an artificial intelligence-based flame detector and a calibration and analysis system thereof. Background Art
[0002] The operating principle of the flame detector is based on detecting the light intensity, flicker frequency and wavelength of the flame. Optical elements such as spectral sensors respond to the ultraviolet, visible light and infrared radiation emitted by the flame, analyze and calculate these characteristics to determine the fire and issue an alarm signal.
[0003] It should be noted that flame detectors are affected by environmental factors such as temperature and humidity, and are prone to failure or false alarms. For example, the detection may not fully cover all possible failure conditions, resulting in the risk of missed detection of the detector. For this, it is necessary to arrange on-site tests regularly, but on-site tests require a lot of manpower and time resources. At the same time, inspectors will also be affected by the safety of the inspection environment and system failures. This application proposes a solution. Summary of the invention
[0004] The purpose of the present invention is to provide an artificial intelligence-based flame detector and a calibration and analysis system thereof. For conventional flame detectors, the optical elements therein may fail or give false alarms due to environmental factors such as temperature / humidity, which requires human and time resources to conduct on-site testing.
[0005] The object of the present invention can be achieved by the following technical solutions: an artificial intelligence-based flame detector, comprising a shell cover, a shell body and a base, wherein a mounting plate and an integrated component are arranged in the shell body from top to bottom, wherein the integrated component comprises an optical sensor, a light source, a signal light and a circuit board, wherein a reflective baffle is installed at a position corresponding to the optical sensor on the mounting plate, wherein the reflective baffle is in a grid shape, and an optical lens is arranged on the shell cover;
[0006] The flame detector includes an active self-checking action and a real-time detection action, and the circuit board is provided with a verification and analysis system associated with the optical sensor, the active self-checking action and the real-time detection action.
[0007] It is further configured that: the verification and analysis system is composed of an information integration unit, an information analysis unit and a parameter correction unit, the information integration unit is used to collect the previous source data and dynamic data, and input the previous source data and dynamic data into the information analysis unit;
[0008] The dynamic data is analyzed in the information analysis unit, and the difference M is calculated by the dynamic data, and M is sent to the parameter correction unit, and the intermediate threshold range of the difference M is set to 0.91<M<1.09;
[0009] In the parameter correction unit, a parameter simulation correction process is performed on the previous source data according to the difference M, an optimized value of the associated dynamic data is obtained in the parameter simulation correction process, and the parameter correction unit is performed on the previous source data with the optimized value.
[0010] It is further configured as follows: the front source data and the dynamic data are respectively applied to the real-time detection action and the active self-check action, and the front source data includes the front light source signal XA and the rear output signal YA; the dynamic data is applied to the active self-check action, and includes the differential light source signal Xi generated by the light source and the differential output signal Yi generated by the optical sensor receiving the differential light source signal; and a self-check cycle is set in the real-time detection action and the active self-check action, and the difference M is calculated as M=1-(Yi-YA) / YA.
[0011] It is further set as follows: In the process of parameter simulation correction, the following contents are included:
[0012] When M>1.09, it indicates that there is a downlink deviation problem in the analysis process performed by the optical sensor, and M-1.09 is taken, and the difference order light source signal in the corresponding value is optimized again to (1+n*M-1.09*n)*Xi, and (1+n*M-1.09*n)*Xi is sent to the light source to regenerate the difference order output signal, and the difference M is calculated again synchronously;
[0013] When M<0.91, it indicates that there is an upward deviation problem in the analysis process performed by the optical sensor. Take M-0.91, optimize the difference-order light source signal in the corresponding value to (1-0.91*n+M*n)*Xi, and calculate the difference M again with (1-0.91*n+M*n)*Xi.
[0014] It is further configured as follows: the parameter model correction process is used to perform the following parameter correction process, the parameter correction process is applied to the real-time detection action, and includes the following parts:
[0015] When M>1.09, the rear-position output signal calculated by the optical sensor is YA*(1+n*M-1.09*n); when M<0.91, the rear-position output signal calculated by the optical sensor is (1-0.91*n+M*n)*YA, and the maximum threshold of the optical sensor is set in the information integration unit, and YA*(1+n*M-1.09*n) or (1-0.91*n+M*n)*YA is compared with the maximum threshold to generate a warning message.
[0016] The present invention has the following beneficial effects:
[0017] 1. The overall structure is a simple improvement on the conventional flame detector. The optical sensor is located inside the structure to prevent the environmental dust in the external environment from affecting the optical sensor. The operation process of the flame detector is divided into real-time detection action and active detection action. The real-time detection action is carried out continuously, and the active detection action is used as a self-check node in the real-time detection. The active detection action is used as the detection process in the overall operation process, which will not affect the normal operation process of the flame detector, and there is no need to use manual detection methods. The detection stage in the overall operation process is completed in an autonomous manner, and no manpower and time are required;
[0018] 2. Based on the above technology, the overall process is to optimize the difference-order light source signal, and to carry out interrelated parameter simulation correction process and parameter correction process respectively. The parameter simulation correction process is the front stage of the parameter correction process, which specifically recalculates the difference therein, and optimizes and improves the subsequent output according to the recalculation process of the difference, so that the parameters of the real-time detection action in the next stage can be corrected, and the alarm parameters can be corrected again when a false alarm or failure may occur. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0020] Figure 1 This is a disassembled diagram of the artificial intelligence-based flame detector proposed by the present invention;
[0021] Figure 2 This is a schematic diagram of the structure of the circuit board in the artificial intelligence-based flame detector proposed by the present invention;
[0022] Figure 3 This is a schematic diagram of the structure of the reflective baffle and the optical sensor in the artificial intelligence-based flame detector proposed by the present invention;
[0023] Figure 4 This is an operating block diagram of the verification and analysis system in the artificial intelligence-based flame detector proposed by the present invention.
[0024] In the figure: 1, shell cover; 2, shell body; 3, mounting plate; 4, integrated component; 5, base; 401, optical sensor; 402, light source; 403, signal light; 404, circuit board; 405, reflective baffle. DETAILED DESCRIPTION
[0025] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than 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.
[0026] Embodiment 1: For conventional flame detectors, the optical elements thereof may fail or give false alarms due to environmental factors such as temperature and humidity, which requires human and time resources to conduct on-site testing. The following technical solutions are proposed:
[0027] Reference Figures 1 to 3 The artificial intelligence-based flame detector in this embodiment includes a shell cover 1, a shell body 2 and a base 5. A mounting plate 3 and an integrated component 4 are arranged in the shell body 2 from top to bottom. The integrated component 4 includes an optical sensor 401, a light source 402, a signal light 403, and a circuit board 404. A reflective baffle 405 is installed at a position corresponding to the optical sensor 401 on the mounting plate 3. The reflective baffle 405 is in a grid shape. An optical lens is arranged on the shell cover 1;
[0028] The flame detector includes an active self-checking action and a real-time detection action. The circuit board 404 is provided with a verification and analysis system for the associated optical sensor 401, the active self-checking action and the real-time detection action.
[0029] Basic principle: The overall structure is basically consistent with the operating principle of a conventional flame detector. The overall structure is arranged in a specified environment, and the ambient light is projected into the optical sensor 401 through the shell cover 1-reflective baffle 405. The optical sensor 401 analyzes and compares the ambient light 1, thereby determining the flame warning signal through the analysis and calculation of the ambient light 1. This part will not be elaborated in detail;
[0030] It should be noted that: with respect to the reflective baffle 405, it is essentially in the shape of a grille. First, it is necessary to ensure that it does not affect the optical sensor 401 from receiving external flame signals. Therefore, after the light source 402 sends out a corresponding light signal, the light signal generated by the light source 402 will also be reflected to the optical sensor at the same time. Therefore, it can be understood that: the active self-detection action is that the light source 402 actively generates a light signal and is received by the optical sensor 401. The light signal generated is mainly a simulation of flame light radiation; and the real-time detection action is the light signal in the external environment and acts on the optical sensor 401 through the reflective baffle 405. The active self-detection action is a self-detection correction of the real-time detection action.
[0031] Embodiment 2: This embodiment is based on the technical solution in Embodiment 1, and explains the active self-checking action and the real-time detection action, and specifically explains the verification and analysis system:
[0032] The calibration and analysis system of the flame detector based on artificial intelligence in this embodiment includes an information integration unit, an information analysis unit and a parameter correction unit. The information integration unit is used to collect front-end source data and dynamic data, and input the front-end source data and dynamic data into the information analysis unit, wherein the front-end source data is applied to the real-time detection action, and the front-end source data includes the front-end light source signal XA and the rear-end output signal YA in the real-time detection action, and the dynamic data is applied to the active self-checking action, and includes the differential light source signal Xi generated by the light source 402 and the differential output signal Yi generated by the optical sensor 401 receiving the differential light source signal, and it should be noted that: wherein the real-time detection action is continuously performed, and a self-checking cycle is set during the continuous process of the real-time detection action, such as the self-checking cycle time is 2h, so that the active self-checking action is started after the real-time detection action is continuously performed for 2h;
[0033] In the information analysis unit, the analysis process mainly focuses on dynamic data, in which the difference-order light source signal Xi is set with reference to multiple groups of previous light source signals XA. It should be noted that the light source signal is a very vague concept, so the light source signal in the present invention is only used as a variable, specifically the output signal is used as a key parameter, so it can be understood that the previous light source signal XA is obtained by feedback from the subsequent output signal YA, and then the relative maximum value XA is extracted from the multiple groups of previous light source signals XA. max and relative minimum value XA min , and the corresponding relative maximum value XA max and relative minimum value XA min The output signal of the latter position is thus at the relative maximum value XA max and relative minimum value XA min Generate the corresponding difference-order light source signal for reference, such as: Xi max and Xi min , the light sources 402 are Xi max and Xi minThe corresponding light source signal is simulated and mapped to the optical sensor 401, so that the corresponding differential output signal is generated in the optical sensor 401, and the differential output signal and the post-position output signal are matched in the information analysis unit, specifically, the differential output signal and the post-position output signal are digitized, which is essentially the electrical signal intensity generated when the optical sensor 401 receives the light signal, so that the difference M is calculated between the differential output signal and the post-position output signal, specifically M=1-(Yi-YA) / YA, when M=1, it is used to indicate that the flame detector is in the best ideal state, and in the case of M≠1, it is necessary to set the intermediate threshold range in the difference to: 0.91<M<1.09, and then it is understood that: when 0.91<M<1.09, it indicates that the flame detector is in normal operating state, on the contrary, when M>1.09 or M<0.91, it indicates that the flame detector is in abnormal operating state.
[0034] Embodiment 3: This embodiment explains the parameter correction unit in combination with the relevant contents in Embodiment 2, specifically, performing parameter simulation correction according to the difference M, including the following contents:
[0035] When M>1.09, Yi<YA, (Yi-YA) / YA is a negative value, indicating that there is a downward deviation in the analysis process performed by the optical sensor 401. In this state, M-1.09 is taken, and the difference-order light source signal in the corresponding value is optimized again to (1+n*M-1.09*n)*Xi, and (1+n*M-1.09*n)*Xi is sent to the light source 402 to regenerate the light signal, and the light signal is used again to act on the optical sensor 401 to regenerate the difference-order output signal, and the difference M is calculated again synchronously;
[0036] When M<0.91, Yi>YA, (Yi-YA) / YA is a positive value, indicating that the analysis process performed by the optical sensor 401 has an upward deviation. In this state, M-0.91 is taken, and the difference order light source signal in the corresponding value is optimized again to (1-0.91*n+M*n)*Xi, and the difference M is calculated again with (1-0.91*n+M*n)*Xi;
[0037] Combining the above two paragraphs, where n is a natural number, based on n=1, until the calculated difference M satisfies 0.91<M<1.09, the warning judgment process of the flame detector is optimized again, specifically including the following states:
[0038] S1: In the real-time detection action, the warning level is determined based on the post-position output signal, specifically based on the output signal generated by the optical sensor 401 after receiving the light source, and the maximum threshold of the output signal of the optical sensor 401 is set in the information integration unit. It can be understood that when YA> the maximum threshold, the flame detector sends out a warning message;
[0039] S2: For the parameter simulation correction process made when M>1.09, in the real-time detection action in the next stage, the parameters of the post-position output signal YA are corrected. Regardless of the different degrees of external ambient light received by the optical sensor 401, the calculated post-position output signal is YA*(1+n*M-1.09*n). On the contrary, when M<0.91, the calculated post-position output signal is (1-0.91*n+M*n)*YA. The optimized YA*(1+n*M-1.09*n) or (1-0.91*n+M*n)*YA is compared with the maximum threshold to determine whether the flame detector needs to issue a warning message.
[0040] To sum up: the operation process of the flame detector is divided into real-time detection action and active detection action. The real-time detection action is carried out continuously, and the active detection action is used as the self-test node in the real-time detection. Its purpose is: the optical sensor is located inside the structure. On the basis of not affecting the normal operation of the flame detector, the corresponding light source is simulated by the light source to radiate on the optical sensor, thereby forming a simulated self-test method. On this technical basis, the differential light source signal is optimized, and the parameter simulation correction and parameter correction process are performed respectively. Among them, the parameter simulation correction process is to recalculate the difference, and the parameter correction process is to optimize and improve the post-output signal. On the basis of false alarm or failure, the alarm parameters are re-corrected.
[0041] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.
[0042] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0043] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.
Claims
1. An artificial intelligence-based flame detector, comprising a housing cover (1), a housing body (2) and a base (5), characterized in that: The housing body (2) is provided with a mounting plate (3) and an integrated component (4) from top to bottom, the integrated component (4) comprising an optical sensor (401), a light source (402), a signal light (403), and a circuit board (404); a reflective baffle (405) is installed at a position on the mounting plate (3) corresponding to the optical sensor (401); the reflective baffle (405) is in a grid shape; and an optical lens is provided on the housing cover (1); The flame detector includes an active self-checking action and a real-time detection action, and the circuit board (404) is provided with a verification and analysis system associated with the optical sensor (401), the active self-checking action and the real-time detection action; The verification and analysis system is composed of an information integration unit, an information analysis unit and a parameter correction unit. The information integration unit is used to collect the previous source data and dynamic data, and input the previous source data and dynamic data into the information analysis unit; The dynamic data is analyzed in the information analysis unit, and the difference M is obtained by calculating the dynamic data, and M is sent to the parameter correction unit, and the intermediate threshold range of the difference M is set to 0.91<M<1.09; in the parameter correction unit, the parameter simulation correction process is performed on the front-end source data according to the difference M, and the optimized value of the associated dynamic data is obtained in the parameter simulation correction process, and the parameter correction unit is performed on the front-end source data with the optimized value; The front source data and the dynamic data are respectively applied to the real-time detection action and the active self-check action, and the front source data includes the front light source signal XA and the rear output signal YA. The dynamic data is applied to the active self-check action and includes the differential light source signal Xi generated by the light source (402) and the differential output signal Yi generated by the optical sensor (401) receiving the differential light source signal. A self-check cycle is set in the real-time detection action and the active self-check action, and the difference M is calculated as M=1-(Yi-YA) / YA. The parameter simulation correction process includes the following: When M>1.09, it indicates that there is a downlink deviation problem in the analysis process performed by the optical sensor (401), and M-1.09 is taken, and the difference order light source signal in the corresponding value is optimized again to (1+n*M-1.09*n)*Xi, and (1+n*M-1.09*n)*Xi is sent to the light source (402) to regenerate the difference order output signal, and the difference M is calculated again synchronously; When M<0.91, it indicates that the analysis process performed by the optical sensor (401) has an upward deviation problem, and M-0.91 is taken, and the difference order light source signal in the corresponding value is optimized again to (1-0.91*n+M*n)*Xi, and the difference M is calculated again with (1-0.91*n+M*n)*Xi; The following parameter correction process is performed with the parameter model correction process. The parameter correction process is applied to the real-time detection action and includes the following parts: When M>1.09, the post-position output signal calculated by the optical sensor (401) is YA*(1+n*M-1.09*n); when M<0.91, the post-position output signal calculated by the optical sensor (401) is (1-0.91*n+M*n)*YA, and a maximum threshold value in the optical sensor (401) is set in the information integration unit, and YA*(1+n*M-1.09*n) or (1-0.91*n+M*n)*YA is compared with the maximum threshold value to generate a warning message.
Citation Information
Patent Citations
Ultra-violet and infrared composite flame detecting alarm having optical self-checking function and method thereof
CN101383075A
Direct-injection type flame detector with self-checking light source and flame detection method
CN102496236A
Self-adaptive optical system based on voltage decoupling controlled multiple wave-front correctors
CN103293663A
Device for testing focal-plane shutters
SU1525668A1
