Artificial intelligence-based gas detection sensor ultra-low power consumption control method and system

Abstract

The application discloses an artificial intelligence-based gas detection sensor ultra-low power consumption control method and system, relates to the technical field of gas detection, and comprises the following steps: obtaining a first-stage heating time based on an optimal temperature and time threshold acquisition method; obtaining a heating time interval and a second-stage heating time based on a normal voltage function, an abnormal voltage function and a time threshold acquisition method; setting a heating mode based on the first-stage heating time, the heating time interval and the second-stage heating time, and marking the heating mode as a low-power-consumption heating mode; and performing gas detection in the low-power-consumption heating mode; and the application is used to solve the problem that in the prior art, intermittent heating time cannot be set based on the sensor itself, and gas detection cannot be performed in real time during intermittent heating, so that the problem that power consumption cannot be reduced and detection effectiveness cannot be maintained simultaneously cannot be solved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of gas detection, in particular to an ultra-low power consumption control method and system for a gas detection sensor based on artificial intelligence. BACKGROUND

[0002] With the development of gas detection sensors, various types of gas sensors are included, including semiconductor gas sensors, which need to heat the resistance when working, and heating will generate a large amount of power consumption.

[0003] Reducing the heating of the resistance in the semiconductor gas sensor can correspondingly reduce the corresponding power consumption, but will affect the normal work of the conventional semiconductor gas sensor, and intermittent heating can also be set, but the interval time of the intermittent heating is subjective, and gas detection cannot be performed when not heating, missing the best opportunity to detect gas, such as the patent application with the application number CN119827723A, which discloses an ultra-low power consumption wireless transmission gas detector, which fails to set the intermittent heating time based on the sensor itself and can perform real-time gas detection during intermittent heating. The prior art fails to set the intermittent heating time based on the sensor itself and can perform real-time gas detection during intermittent heating, resulting in the inability to simultaneously achieve power reduction and detection effectiveness. SUMMARY

[0004] The application aims to at least solve one of the technical problems in the prior art, by obtaining the first stage heating time based on the optimal temperature and time threshold value acquisition method, obtaining the heating time interval and the second stage heating time based on the normal voltage function, the abnormal voltage function and the time threshold value acquisition method, setting the heating mode based on the first stage heating time, the heating time interval and the second stage heating time, and marking it as a low-power heating mode, and performing gas detection in the low-power heating mode, to solve the problem that the prior art fails to set the intermittent heating time based on the sensor itself and can perform real-time gas detection during intermittent heating, resulting in the inability to simultaneously achieve power reduction and detection effectiveness.

[0005] To achieve the above-mentioned purpose, the application provides an ultra-low power consumption control method for a gas detection sensor based on artificial intelligence, which comprises the following steps:

[0006] Obtain the optimal detection temperature of the resistance when the sensor works, which is marked as the optimal temperature;

[0007] Obtain the first stage heating time based on the optimal temperature and time threshold value acquisition method;

[0008] When no detection gas is contained, a function of the voltage across the resistance with respect to time after the resistance is heated to the optimal temperature and then the heating is stopped is obtained, and is marked as a normal voltage function;

[0009] When the detection gas is contained, a function of the voltage across the resistance with respect to time after the resistance is heated to the optimal temperature and then the heating is stopped is obtained, and is marked as an abnormal voltage function;

[0010] The heating time interval and the second stage heating time are obtained based on the normal voltage function, the abnormal voltage function and the time threshold value obtaining method;

[0011] The heating mode is set based on the first stage heating time, the heating time interval and the second stage heating time, and is marked as a low-power-consumption heating mode;

[0012] The gas detection is performed in the low-power-consumption heating mode.

[0013] Further, the first stage heating time is obtained based on the optimal temperature and the time threshold value obtaining method, and includes the following sub-steps:

[0014] A heating time of the resistance heated to the optimal temperature in the current environment is obtained, and is marked as a first heating time;

[0015] The first heating time is taken as an experimental time, and the time threshold value is obtained by using the time threshold value obtaining method, and is marked as the first stage heating time;

[0016] The time threshold value obtaining method includes:

[0017] A first number of experimental times is obtained;

[0018] The experimental times are sorted in ascending order and are marked as Hs1 to Hs i ;

[0019] A smaller position ratio is calculated as Ji=a1*S1, where Ji is the smaller position ratio, a1 is a smaller coefficient, the range of a1 is (0, 0.5), and S1 is the first number;

[0020] If the smaller position ratio is an integer, Hs （Ji） is marked as a smaller position value; if the smaller position ratio is not an integer, integers on the left and right sides of Ji are obtained, and are marked as Jiz and Jiy respectively, and a mean value of Hs （Jiz） and Hs （Jiy） is calculated, and is marked as the smaller position value;

[0021] It is determined whether 0.5*S1 is an integer, if yes, Hs （0.5*S1） is marked as a middle position value; if not, integers on the left and right sides of 0.5*S1 are obtained, and are marked as Jzz and Jzy respectively, and a mean value of Hs （Jzz） and Hs （Jzz）The average of the Hs

[0022] The larger coefficient is calculated as: a2=[(1 / a1)-1]*a1; wherein a2 is the larger coefficient;

[0023] The larger position ratio is calculated as: Jx=a2*S1, wherein Jx is the smaller position ratio;

[0024] If the larger position ratio is an integer, Hs （Jx） If the larger position ratio is not an integer, the integers on the left and right sides of Jx are obtained, respectively marked as Jxz and Jxy, and Hs （Jxz） is calculated as: Hs （Jxy） The average of the Hs

[0025] The time threshold is calculated as: Sy=Wz+b1*(Wx-Wi) / 2; wherein Sy is the time threshold, Wx is the larger position value, Wi is the smaller position value, Wz is the intermediate position value, and b1 is a threshold adjustment coefficient.

[0026] Further, when there is no detected gas, a function of the voltage across the resistance with respect to time after heating to the optimal temperature and stopping heating is obtained, and is marked as a normal voltage function, including the following sub-steps:

[0027] A second number of times of stopping heating at the optimal temperature and starting timing is obtained, and the timing time at this time is marked as a first timing time. The voltage across the resistance at this time is obtained every interval T, and is marked as a normal voltage.

[0028] The normal voltage at each same interval is marked as a normal interval voltage.

[0029] The range of the normal interval voltage is evenly divided into c1 equal range intervals, and is marked as a normal voltage division interval.

[0030] The frequency of each normal voltage division interval is counted, and is marked as a normal division frequency.

[0031] A histogram is drawn with the normal interval voltage as the X-axis, the normal division frequency as the Y-axis, and the normal voltage division interval as the histogram interval, and is marked as a normal voltage histogram.

[0032] The second number is marked as S2.

[0033] The first smaller frequency threshold is calculated as: F1=d1*S2 / c1; wherein F1 is the first division frequency threshold, and d1 is a first smaller frequency proportion.

[0034] The normal division frequencies less than or equal to the first smaller frequency threshold are marked as a first smaller frequency.

[0035] Mark the normal partition frequency number at the leftmost and rightmost of the normal power consumption histogram as normal edge frequency number;

[0036] Determine whether the normal edge frequency number is the first smaller frequency number, if yes, delete the image part of the normal edge frequency number being the first smaller frequency number in the normal voltage histogram, repeat the above operation on the normal voltage histogram after deletion until it is not, and then stop; mark the normal voltage histogram after stopping as a screening normal voltage histogram;

[0037] Obtain the maximum and minimum values of the normal partition power consumption frequency abscissa in the screening normal voltage histogram, and mark them as the first screening threshold and the second screening threshold respectively.

[0038] Further, when there is no detection gas, obtain the function of the voltage across the resistor changing with time after heating to the optimal temperature and stopping heating, and mark it as a normal voltage function, which further includes the following sub-steps:

[0039] Calculate the mean value of the normal interval voltage between the first screening threshold and the second screening threshold, and mark it as a proposed normal interval voltage;

[0040] Obtain all the proposed normal interval voltages;

[0041] Establish a plane rectangular coordinate system with time as the X-axis data and voltage as the Y-axis data, and mark it as a voltage coordinate system;

[0042] Take each proposed normal interval voltage and the corresponding first time as the ordinate and abscissa of the normal voltage coordinate point respectively, and plot all the normal voltage coordinate points in the normal voltage coordinate system;

[0043] Obtain the normal voltage function by polynomial fitting of all the normal voltage coordinate points.

[0044] Further, when there is detection gas, obtain the function of the voltage across the resistor changing with time after heating to the optimal temperature and stopping heating, and mark it as an abnormal voltage function, which includes the following sub-steps:

[0045] Obtain the third number of times of stopping heating at the optimal temperature to start timing, mark the timing time at this time as the second timing time, release the detection gas at different second timing times, and obtain the voltage across the resistor at this time, which is marked as an abnormal voltage;

[0046] Mark the abnormal voltage at each second timing time as an abnormal interval voltage;

[0047] Divide the range of abnormal interval voltages into c2 equal range intervals on average, and mark them as abnormal voltage division intervals;

[0048] Count the frequency of each abnormal voltage interval, and mark it as an abnormal division frequency;

[0049] Draw a histogram with the abnormal interval voltage as the X-axis, the abnormal division frequency as the Y-axis, and the abnormal voltage interval as the histogram interval, and mark it as the abnormal voltage histogram;

[0050] Mark the third number as S3;

[0051] Calculate the second smaller frequency threshold value as: F2=d2*S3 / c2; wherein F2 is the second division frequency threshold value, d2 is the second smaller frequency proportion;

[0052] Mark the abnormal division frequency less than or equal to the second smaller frequency threshold value as the second smaller frequency;

[0053] Mark the abnormal division frequency at the leftmost and rightmost sides of the abnormal power consumption histogram as the abnormal side frequency;

[0054] Determine whether the abnormal side frequency is the second smaller frequency, if so, delete the image part of the abnormal side frequency as the second smaller frequency in the abnormal voltage histogram, and repeat the above operation on the deleted abnormal voltage histogram until it is not stopped; mark the stopped abnormal voltage histogram as the screened abnormal voltage histogram;

[0055] Obtain the maximum and minimum values of the abnormal division power frequency abscissa in the screened abnormal voltage histogram, and mark them as the third screening threshold value and the fourth screening threshold value.

[0056] Further, when containing the detection gas, obtain the relationship function of the resistance voltage change with time after heating to the optimal temperature and stopping heating, and mark it as the abnormal voltage function, which further includes the following sub-steps:

[0057] Calculate the mean value of the abnormal interval voltage between the third screening threshold value and the fourth screening threshold value, and mark it as the proposed abnormal interval voltage;

[0058] Obtain the proposed abnormal interval voltage of all abnormal interval voltages;

[0059] Take each proposed abnormal interval voltage and the corresponding second timing time as the ordinate and abscissa of the abnormal voltage coordinate point respectively, and draw all abnormal voltage coordinate points in the voltage coordinate system;

[0060] Polynomial fitting is performed on all abnormal voltage coordinate points to obtain the abnormal voltage function.

[0061] Further, based on the normal voltage function, the abnormal voltage function, and the time threshold value obtaining method, the heating time interval and the second stage heating time are obtained, which includes the following sub-steps:

[0062] Calculate the difference between all the first screening threshold and the second screening threshold, marked as threshold difference;

[0063] Obtain the mean value of the absolute value of the threshold difference, marked as range difference threshold;

[0064] Mark half of the range difference threshold as the resolution threshold;

[0065] In the voltage coordinate system, mark the difference between the ordinate value of the abnormal voltage function and the ordinate value of the normal voltage function at the same time as the abnormal difference;

[0066] When the abnormal difference is equal to the resolution difference, the corresponding abscissa minimum value in the voltage coordinate system is marked as the heating time interval;

[0067] After the heating time interval, the heating time of the resistance to the optimal temperature is marked as the second heating time;

[0068] Take the second heating time as the experimental time to obtain the time threshold using the time threshold obtaining method, and mark it as the second stage heating time.

[0069] Further, based on the first stage heating time, the heating time interval and the second stage heating time, set the heating mode, marked as low-power heating mode, including the following sub-steps:

[0070] Set the gas detection sensor to start heating the resistance when it is working, stop heating after the first stage heating time, start heating after the heating time interval, and stop heating after the second stage heating time. Then repeat the steps of starting heating after the heating time interval and stopping heating after the second stage heating time.

[0071] Further, the low-power heating mode is used for gas detection, including the following steps:

[0072] In the low-power heating mode and without containing the detection gas, mark the voltage on both sides of the resistance as the detection voltage; mark the timing time started by the low-power heating mode as the working time;

[0073] Take the working time as the X-axis data and the detection voltage as the Y-axis data to establish a plane rectangular coordinate system, marked as the detection coordinate system;

[0074] Take the detection voltage and the corresponding working time as the ordinate and abscissa of the detection coordinate point respectively, and draw all the detection coordinate points in the detection coordinate system;

[0075] Obtain the detection voltage function by polynomial fitting of all the detection coordinate points;

[0076] Obtain the working time and the detection voltage when the gas detection sensor is working, and mark them as real-time time and real-time voltage respectively;

[0077] The detection voltage obtained by substituting the real-time time into the detection voltage function is marked as a predicted normal voltage;

[0078] The difference between the real-time voltage and the predicted normal voltage is calculated and marked as a detection difference value;

[0079] If the detection difference value is greater than a resolution threshold value, a detection gas signal is sent out.

[0080] The application also provides an ultra-low power consumption control system for a gas detection sensor based on artificial intelligence, comprising a temperature acquisition module, a time acquisition module, a first function acquisition module, a second function acquisition module, a heating interval acquisition module, a heating setting module, and a gas detection module.

[0081] The temperature acquisition module is used to acquire an optimal detection temperature of the resistance of the sensor when working, which is marked as an optimal temperature.

[0082] The time acquisition module is used to acquire a first-stage heating time based on an optimal temperature and a time threshold value acquisition method.

[0083] The first function acquisition module is used to acquire a relationship function of the voltage across the resistance changing with time after heating to the optimal temperature and stopping heating when there is no detection gas, which is marked as a normal voltage function.

[0084] The second function acquisition module is used to acquire a relationship function of the voltage across the resistance changing with time after heating to the optimal temperature and stopping heating when there is detection gas, which is marked as an abnormal voltage function.

[0085] The heating interval acquisition module is used to acquire a heating time interval and a second-stage heating time based on the normal voltage function, the abnormal voltage function, and a time threshold value acquisition method.

[0086] The heating setting module is used to set a heating mode based on the first-stage heating time, the heating time interval, and the second-stage heating time, which is marked as a low-power-consumption heating mode.

[0087] The gas detection module is used to perform gas detection in the low-power-consumption heating mode.

[0088] The application has the following advantages: the first-stage heating time is acquired based on an optimal temperature and a time threshold value acquisition method, the heating time interval and the second-stage heating time are acquired based on a normal voltage function, an abnormal voltage function, and a time threshold value acquisition method, the heating mode is set based on the first-stage heating time, the heating time interval, and the second-stage heating time, which is marked as a low-power-consumption heating mode, and gas detection is performed in the low-power-consumption heating mode, so that the heating mode can be set based on the sensor itself, and normal detection can be performed while reducing power consumption.

[0089] The time threshold acquisition method has the advantages that the data can be further screened, the function fitted subsequently is more accurate, and the accuracy of subsequent gas detection is improved. BRIEF DESCRIPTION OF DRAWINGS

[0090] Figure 1 A principle block diagram of the system of the present application;

[0091] Figure 2 A schematic diagram of the normal voltage histogram of the present application;

[0092] Figure 3 A schematic diagram of the normal voltage histogram of the present application;

[0093] Figure 4 A schematic diagram of the normal voltage function of the present application;

[0094] Figure 5 A schematic diagram of the abnormal voltage function of the present application;

[0095] Figure 6 A schematic diagram of the detection voltage function of the present application;

[0096] Figure 7 A schematic diagram of the method of the present application. DETAILED DESCRIPTION

[0097] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0098] Embodiment 1, please refer to Figure 1 As shown in the figure, the present application provides an ultra-low power control system for a gas detection sensor based on artificial intelligence, which comprises a temperature acquisition module, a time acquisition module, a first function acquisition module, a second function acquisition module, a heating interval acquisition module, a heating setting module, and a gas detection module.

[0099] The temperature acquisition module is used to acquire the optimal detection temperature of the resistance of the sensor when working, which is marked as the optimal temperature.

[0100] The time acquisition module is used to acquire the first-stage heating time based on the optimal temperature and a time threshold acquisition method.

[0101] The time acquisition module is configured with a time acquisition strategy, and the time acquisition strategy comprises:

[0102] Obtaining the heating time of resistance heating to the optimal temperature in the current environment, marked as the first heating time;

[0103] The first heating time is used as the experimental time to obtain the time threshold value by the time threshold value obtaining method, marked as the first stage heating time;

[0104] The time threshold value obtaining method comprises:

[0105] Obtaining a first quantity of experimental times;

[0106] The experimental times are sorted in ascending order and marked as Hs1 to Hs i ;

[0107] The smaller position ratio is Ji=a1*S1, wherein Ji is the smaller position ratio, a1 is a smaller coefficient, the range of a1 is (0, 0.5), S1 is the first quantity; in order to select the smaller position value distributed in the smaller half area, a1 is set between 0 and 0.5, and an intermediate value is selected to ensure that the selected first position value has the representativeness of the smaller half area, and the intermediate value is preferably 0.25;

[0108] If the smaller position ratio is an integer, Hs （Ji） is marked as the smaller position value; if the smaller position ratio is not an integer, the integers on the left and right sides of Ji are obtained and marked as Jiz and Jiy respectively, and the mean value of Hs （Jiz） and Hs （Jiy） is obtained and marked as the smaller position value;

[0109] It is judged whether 0.5*S1 is an integer, if yes, Hs （0.5*S1） is marked as the intermediate position value; if not, the integers on the left and right sides of 0.5*S1 are obtained and marked as Jzz and Jzy respectively, and the mean value of Hs （Jzz） and Hs （Jzz） is obtained and marked as the intermediate position value;

[0110] The larger coefficient is a2=[(1 / a1)-1]*a1; wherein a2 is the larger coefficient; in order to select the larger position value distributed in the larger half area, a2 is set between 0.5 and 1, and in order to ensure that a1 to the intermediate value and a2 to the intermediate value have equal proportions, a2=0.75;

[0111] The larger position ratio is Jx=a2*S1, wherein Jx is the smaller position ratio;

[0112] If the larger position ratio is an integer, Hs （Jx） is marked as the larger position value; if the larger position ratio is not an integer, the integers on the left and right sides of Jx are obtained and marked as Jxz and Jxy respectively, and the mean value of Hs （Jxz） and Hs（Jxy） the average of the values of the smaller position values;

[0113] The time threshold is calculated as Sy = Wz + b1*(Wx-Wi) / 2; where Sy is the time threshold, Wx is the larger position value, Wi is the smaller position value, Wz is the intermediate position value, and b1 is the threshold adjustment coefficient; where the setting of b1 is based on the distribution of the data, the more concentrated the data, the smaller b1, because under the same conditions, the time spent should be basically the same, so the data is more concentrated, and because a1 is set to 0.25, so when the data is uniformly distributed, b1 is 2 when Sy reaches the maximum value, and b1 is 1 when Sy reaches a value similar to Wx, therefore b1 is a value between 1 and 2, and a middle value between 1 and 2 is selected as a representative, b1 is 1.5;

[0114] In actual application, for example, when the first number is 20, a1 is 0.25, a2 is 0.75, and b1 is 1.5, the smaller position ratio is calculated as Ji = 0.25*20 = 8, which is an integer, and Hs （8） The corresponding 6.16 min is marked as the smaller position value; determine whether 0.5*20 is an integer, 10 is an integer, and Hs （10） The corresponding 6.25 min is marked as the intermediate position value, and the larger position ratio is calculated as Ji = 0.75*20 = 16, which is an integer, and Hs （16） The corresponding 6.34 min is marked as the larger position value; the time threshold is calculated as Sy = 6.25 + 1.5*(6.34-6.16) / 2 = 6.385 min.

[0115] The first function acquisition module is configured with a first screening threshold and a second screening threshold acquisition strategy, and the first screening threshold and the second screening threshold acquisition strategy include:

[0116] The first function acquisition module is configured with a first screening threshold and a second screening threshold acquisition strategy, and the first screening threshold and the second screening threshold acquisition strategy include:

[0117] The second number of times is obtained at the optimal temperature when the heating is stopped, and the timing time at this time is marked as the first timing time. Every interval T, the voltage across the resistor at this time is obtained and marked as the normal voltage. Here, data is obtained every interval T, which can save data processing. A constant current measurement method is used, and when gas is detected, the resistance will change. Only the change in voltage is detected to detect gas;

[0118] The normal voltage at each same interval is marked as the normal interval voltage;

[0119] The range of the normal interval voltage is evenly divided into c1 equal range intervals, which are marked as the normal voltage division intervals;

[0120] The frequency of each normal voltage division interval is counted and marked as a normal division frequency;

[0121] A histogram is drawn with the normal interval voltage as the X-axis, the normal division frequency as the Y-axis, and the normal voltage division interval as the histogram interval, and marked as a normal voltage histogram;

[0122] The second number is marked as S2;

[0123] The first smaller frequency threshold is calculated as: F1=d1*S2 / c1; where F1 is the first division frequency threshold, and d1 is the first smaller frequency proportion. The first smaller frequency threshold means that the data with a small proportion is identified as a smaller part of S2 after equal division, and d1 is set to 0.1.

[0124] The normal division frequency less than or equal to the first smaller frequency threshold is marked as the first smaller frequency;

[0125] The normal division frequency at the leftmost and rightmost sides of the normal power consumption histogram is marked as the normal side frequency;

[0126] It is determined whether the normal side frequency is the first smaller frequency. If so, the image part of the normal side frequency as the first smaller frequency in the normal voltage histogram is deleted, and the above operation is repeated on the normal voltage histogram after deletion until it is not. The normal voltage histogram after stopping is marked as a screened normal voltage histogram. This method can screen the data to make the obtained voltage range more accurate;

[0127] The maximum and minimum values of the normal division power frequency abscissa in the screened normal voltage histogram are obtained and marked as the first screening threshold and the second screening threshold, respectively;

[0128] In actual application, please refer to Figure 2 and Figure 3 Taking multiple normal voltages obtained at the same time as an example, the range of normal interval voltage is evenly divided into 7 equal range intervals. The second number is 180, the first smaller frequency threshold is calculated as: F1=0.1*180 / 7=2.6, and the calculation result is rounded to one decimal place. For example, the normal side frequency is 2, and 2 is less than 2.6, which is the first smaller frequency. Therefore, the part of 2 in the normal voltage histogram is deleted. Please refer to Figure 3 , the first screening threshold and the second screening threshold are obtained as 3.12 and 3.22.

[0129] The first function acquisition module is configured with a first function acquisition strategy, and the first function acquisition strategy includes:

[0130] Calculate the mean of the normal interval voltages between the first screening threshold and the second screening threshold, marked as the proposed normal interval voltage;

[0131] Obtain all the proposed normal interval voltages;

[0132] Establish a plane rectangular coordinate system with time as the X-axis data and voltage as the Y-axis data, marked as the voltage coordinate system;

[0133] Draw all normal voltage coordinate points in the normal voltage coordinate system by taking each proposed normal interval voltage and the corresponding first time as the ordinate and abscissa of the normal voltage coordinate point, respectively;

[0134] Obtain the normal voltage function by polynomial fitting of all normal voltage coordinate points;

[0135] In practical applications, please participate Figure 4 Obtain the normal voltage function.

[0136] The second function acquisition module is configured to obtain a relationship function between the voltage across the resistor and time after heating to the optimal temperature and stopping heating when containing a detection gas, marked as an abnormal voltage function;

[0137] The second function acquisition module is configured with a third screening threshold and a fourth screening threshold acquisition strategy, and the third screening threshold and the fourth screening threshold acquisition strategy includes:

[0138] Obtain a third number of times of stopping heating at the optimal temperature and starting timing, mark the timing time at this time as the second timing time, release the detection gas at different second timing times, and obtain the voltage across the resistor at this time, marked as abnormal voltage;

[0139] Mark the abnormal voltage under each second timing time as an abnormal interval voltage;

[0140] Divide the range of abnormal interval voltages into c2 equal range intervals, marked as abnormal voltage division intervals;

[0141] Statistically count the frequency of each abnormal voltage division interval, marked as abnormal division frequency;

[0142] Draw a histogram with the abnormal interval voltage as the X-axis, the abnormal division frequency as the Y-axis, and the abnormal voltage division interval as the histogram interval, marked as the abnormal voltage histogram;

[0143] Mark the third number as S3;

[0144] Calculate the second smaller frequency threshold as: F2=d2*S3 / c2; Wherein F2 is the second division frequency threshold, and d2 is the second smaller frequency ratio;

[0145] mark the abnormal division frequency less than or equal to the second smaller frequency threshold as a second smaller frequency;

[0146] mark the abnormal division frequency at the leftmost and rightmost sides of the abnormal power consumption histogram as an abnormal side frequency;

[0147] determine whether the abnormal side frequency is the second smaller frequency, if yes, delete the image part of the abnormal voltage histogram where the abnormal side frequency is the second smaller frequency, and repeatedly perform the above operation on the abnormal voltage histogram after deletion until it stops; mark the abnormal voltage histogram after stopping as a screened abnormal voltage histogram;

[0148] obtain the maximum and minimum values of the abnormal division power consumption frequency abscissa in the screened abnormal voltage histogram, and mark them as a third screening threshold and a fourth screening threshold respectively; the method for obtaining the third screening threshold and the fourth screening threshold is consistent with the method for obtaining the first screening threshold and the second screening threshold.

[0149] The second function obtaining module is configured with a second function obtaining strategy, and the second function obtaining strategy includes:

[0150] calculate the mean value of the abnormal interval voltage between the third screening threshold and the fourth screening threshold, and mark it as a tentative abnormal interval voltage;

[0151] obtain all tentative abnormal interval voltages;

[0152] take each tentative abnormal interval voltage and the corresponding second timing time as the ordinate and abscissa of the abnormal voltage coordinate point respectively, and draw all abnormal voltage coordinate points in the voltage coordinate system;

[0153] obtain the abnormal voltage function by polynomial fitting of all abnormal voltage coordinate points;

[0154] In practical application, please participate Figure 5 As shown in the figure, the obtained abnormal voltage function, the resistance material and the detection gas will react to increase the resistance at high temperature, so under the condition of constant current, the voltage increases compared with the normal voltage function, but with the passage of time, the temperature decreases, which reduces the effect.

[0155] The heating interval obtaining module is configured with a heating interval obtaining strategy, and the heating interval obtaining strategy includes:

[0156] The heating interval obtaining module is configured with a heating interval obtaining strategy, and the heating interval obtaining strategy includes:

[0157] calculate the difference between all first screening thresholds and second screening thresholds, and mark it as a threshold difference;

[0158] The mean value of the threshold difference absolute value is marked as a range difference threshold value;

[0159] Half of the range difference threshold value is marked as a resolution threshold value; a normal voltage distribution threshold value is obtained, and if it is exceeded, a detection gas can be detected to change the voltage;

[0160] In the voltage coordinate system, the difference between the ordinate value of the abnormal voltage function and the ordinate value of the normal voltage function at the same time is marked as an abnormal difference value;

[0161] When the abnormal difference value is equal to the resolution difference value, the minimum value of the corresponding abscissa in the voltage coordinate system is marked as a heating time interval; the longest stop heating time can be obtained, and the voltage judgment of the detection gas cannot be affected;

[0162] The heating time of the resistance to the optimal temperature after the heating time interval is marked as a second heating time;

[0163] The second heating time is used as the experimental time to obtain the time threshold value by using the time threshold value obtaining method, and is marked as a second stage heating time, so that the accurate second stage heating time can be obtained.

[0164] The heating setting module is configured to set a heating mode based on the first stage heating time, the heating time interval, and the second stage heating time, and is marked as a low-power heating mode;

[0165] In actual application, for example, the resolution difference value 0.1V is obtained, please refer to Figure 5 As shown in the figure, the abnormal difference value gradually decreases with time, and when 6.1 min, the abnormal difference value is less than 0.1V, so the heating time interval is 6.1 min, and the second stage heating time is 5.8 min.

[0166] The heating setting module is configured with a heating setting strategy, and the heating setting strategy includes:

[0167] The gas detection sensor is set to start heating the resistance when it works, stop heating after the first stage heating time, start heating after the heating time interval, stop heating after the second stage heating time, and then repeat the start heating after the heating time interval and the stop heating after the second stage heating time; This method saves the heating power consumption to the maximum extent;

[0168] In actual application, the first stage heating time obtained is 7.2 min, and the gas detection sensor is set to start heating the resistance when it works, stop heating after 7.2 min, start heating after 6.1 min, stop heating after 5.8 min, and then repeat the start heating after 6.1 min and the stop heating after 5.8 min; This method saves the heating power consumption to the maximum extent.

[0169] The gas detection module is used for gas detection in a low-power heating mode;

[0170] The gas detection module is configured with a gas detection strategy, and the gas detection strategy includes:

[0171] In the low-power heating mode and without the detection gas, the voltage on both sides of the resistor is marked as the detection voltage; and the timing time starting from the low-power heating mode is marked as the working time.

[0172] Taking the working time as the X-axis data and the detection voltage as the Y-axis data, a plane rectangular coordinate system is established and marked as a detection coordinate system;

[0173] The detection voltage and the corresponding working time are taken as the ordinate and the abscissa of the detection coordinate point respectively, and all the detection coordinate points are plotted in the detection coordinate system;

[0174] All the detection coordinate points are polynomially fitted to obtain a detection voltage function;

[0175] The working time and the detection voltage of the gas detection sensor during work are obtained and marked as the real-time time and the real-time voltage respectively;

[0176] The detection voltage obtained by substituting the real-time time into the detection voltage function is marked as the predicted normal voltage;

[0177] The difference between the real-time voltage and the predicted normal voltage is calculated and marked as a detection difference;

[0178] If the detection difference is greater than a resolution threshold, a detection gas signal is sent; the resolution threshold normally distinguishes a roughly normal upward amplitude range, and an amplitude exceeding this range can be determined as a voltage change caused by the detection gas;

[0179] In actual application, please participate Figure 6 It should be noted that, after the heating time interval, the heating is started again, and after the second stage heating time, the heating is stopped again; this method saves the maximum heating power consumption, so the function image is basically the same, and this function image can be used to judge the subsequent function changes, for example, at 22.1 min, 22.1-6.1-5.8=10.2 min, which actually corresponds to 10.2 min, and the detection voltage function is substituted into 2.2v, and the real-time voltage is obtained as 2.6v, and the detection difference is 0.4v, which is greater than the resolution threshold of 0.1v, and a detection gas signal is sent.

[0180] Embodiment 2, please refer to Figure 7 As shown in the figure, the application provides an ultra-low power control method for a gas detection sensor based on artificial intelligence, which includes the following steps:

[0181] Step S1, obtaining the optimal detection temperature of the resistance of the sensor when working, marked as the optimal temperature;

[0182] Step S2, obtaining the first-stage heating time based on the optimal temperature and the time threshold obtaining method; Step S2 includes the following sub-steps:

[0183] Step S201, obtaining the heating time of the resistance to the optimal temperature in the current environment, marked as the first heating time;

[0184] Step S202, obtaining the time threshold by using the time threshold obtaining method with the first heating time as the experimental time, marked as the first-stage heating time; Step S202 includes the following sub-steps:

[0185] Step S20201, obtaining the first quantity of experimental times;

[0186] Step S20202, sorting the experimental times in ascending order and marking them as Hs1 to Hs i ;

[0187] Step S20203, obtaining the smaller position ratio as Ji=a1*S1, wherein Ji is the smaller position ratio, a1 is the smaller coefficient, the range of a1 is (0, 0.5), and S1 is the first quantity;

[0188] Step S20204, if the smaller position ratio is an integer, marking Hs （Ji） as the smaller position value; if the smaller position ratio is not an integer, obtaining the integers on the left and right sides of Ji, respectively marked as Jiz and Jiy, and obtaining the mean value of Hs （Jiz） and Hs （Jiy） , marked as the smaller position value;

[0189] Step S20205, determining whether 0.5*S1 is an integer, if yes, marking Hs （0.5*S1） as the middle position value; if not, obtaining the integers on the left and right sides of 0.5*S1, respectively marked as Jzz and Jzy, and obtaining the mean value of Hs （Jzz） and Hs （Jzz） , marked as the middle position value;

[0190] Step S20206, obtaining the larger coefficient as a2=[(1 / a1)-1]*a1; wherein a2 is the larger coefficient;

[0191] Step S20207, obtaining the larger position ratio as Jx=a2*S1, wherein Jx is the smaller position ratio;

[0192] Step S20208, if the larger position ratio is an integer, marking Hs （Jx）The larger position value is marked; if the larger position ratio is not an integer, the integers on the left and right sides of Jx are obtained and marked as Jxz and Jxy, respectively, and Hs is calculated （Jxz） The average of Hs （Jxy） The smaller position value is marked;

[0193] Step S20209, the time threshold is calculated as Sy=Wz+b1*(Wx-Wi) / 2; where Sy is the time threshold, Wx is the larger position value, Wi is the smaller position value, Wz is the intermediate position value, and b1 is the threshold adjustment coefficient.

[0194] Step S3, when there is no detected gas, a function of the voltage across the resistor with respect to time after heating to the optimal temperature and stopping heating is obtained, and is marked as a normal voltage function; step S3 includes the following sub-steps:

[0195] Step S301, a second number of times of starting timing when stopping heating at the optimal temperature is obtained, and the timing time at this time is marked as a first timing time; every T time interval, the voltage across the resistor at this time is obtained and marked as a normal voltage;

[0196] Step S302, the normal voltage at each same interval is marked as a normal interval voltage;

[0197] Step S303, the range of normal interval voltages is evenly divided into c1 equal range intervals, and is marked as a normal voltage division interval;

[0198] Step S304, the frequency of each normal voltage division interval is counted and marked as a normal division frequency;

[0199] Step S305, a histogram is drawn with the normal interval voltage as the X-axis, the normal division frequency as the Y-axis, and the normal voltage division interval as the histogram interval, and is marked as a normal voltage histogram;

[0200] Step S306, the second number is marked as S2;

[0201] Step S307, the first smaller frequency threshold is calculated as F1=d1*S2 / c1; where F1 is the first division frequency threshold, and d1 is the first smaller frequency proportion;

[0202] Step S308, the normal division frequencies less than or equal to the first smaller frequency threshold are marked as the first smaller frequency;

[0203] Step S309, the normal division frequencies at the leftmost and rightmost sides of the normal power consumption histogram are marked as normal side frequencies;

[0204] Step S310, judging whether the normal edge side frequency is the first smaller frequency, if yes, deleting the image part of the normal voltage histogram in which the normal edge side frequency is the first smaller frequency, repeating the above operation on the normal voltage histogram after deletion until it is not, and then stopping; marking the normal voltage histogram after stopping as a screening normal voltage histogram;

[0205] Step S311, obtaining the maximum value and the minimum value of the normal division power consumption frequency abscissa in the screening normal voltage histogram, and marking them as the first screening threshold and the second screening threshold respectively;

[0206] Step S312, calculating the mean value of the normal interval voltage between the first screening threshold and the second screening threshold, and marking it as a tentative normal interval voltage;

[0207] Step S313, obtaining the tentative normal interval voltage of all normal interval voltages;

[0208] Step S314, establishing a plane rectangular coordinate system with time as the X-axis data and voltage as the Y-axis data, and marking it as a voltage coordinate system;

[0209] Step S315, taking each tentative normal interval voltage and the corresponding first time as the ordinate and the abscissa of the normal voltage coordinate point respectively, and drawing all normal voltage coordinate points in the normal voltage coordinate system;

[0210] Step S316, performing polynomial fitting on all normal voltage coordinate points to obtain a normal voltage function.

[0211] Step S4, when containing the detection gas, obtaining a relationship function of the voltage across the resistance varying with time after heating to the optimal temperature and stopping heating, and marking it as an abnormal voltage function; step S4 includes the following sub-steps:

[0212] Step S401, obtaining a third number of times of starting timing after stopping heating at the optimal temperature, marking the timing time at this time as a second timing time, releasing the detection gas at different second timing times, and obtaining the voltage across the resistance at this time, and marking it as an abnormal voltage;

[0213] Step S402, marking the abnormal voltage at each second timing time as an abnormal interval voltage;

[0214] Step S403, evenly dividing the range of the abnormal interval voltage into c2 equal range intervals, and marking them as abnormal voltage division intervals;

[0215] Step S404, counting the frequency of each abnormal voltage division interval, and marking it as an abnormal division frequency;

[0216] Step S405, a histogram is drawn with the abnormal interval voltage as the X-axis, the abnormal division frequency as the Y-axis, and the abnormal voltage division interval as the histogram interval, and is marked as an abnormal voltage histogram;

[0217] Step S406, the third number is marked as S3;

[0218] Step S407, the second smaller frequency threshold is calculated as: F2=d2*S3 / c2; wherein F2 is the second division frequency threshold, and d2 is the second smaller frequency proportion;

[0219] Step S408, the abnormal division frequency less than or equal to the second smaller frequency threshold is marked as the second smaller frequency;

[0220] Step S409, the abnormal division frequency at the leftmost and rightmost of the abnormal power consumption histogram is marked as the abnormal side frequency;

[0221] Step S410, it is judged whether the abnormal side frequency is the second smaller frequency, if yes, the image part of the abnormal side frequency being the second smaller frequency in the abnormal voltage histogram is deleted, and the above operation is repeated on the deleted abnormal voltage histogram until it is not stopped; the abnormal voltage histogram after stopping is marked as a screened abnormal voltage histogram;

[0222] Step S411, the maximum value and the minimum value of the abnormal division power consumption frequency abscissa in the screened abnormal voltage histogram are obtained, and are marked as a third screening threshold and a fourth screening threshold, respectively;

[0223] Step S412, the mean value of the abnormal interval voltage between the third screening threshold and the fourth screening threshold is calculated, and is marked as a proposed abnormal interval voltage;

[0224] Step S413, the proposed abnormal interval voltage of all abnormal interval voltages is obtained;

[0225] Step S414, each proposed abnormal interval voltage and the corresponding second timing time are taken as the ordinate and the abscissa of the abnormal voltage coordinate point, respectively, and all abnormal voltage coordinate points are drawn in the voltage coordinate system;

[0226] Step S415, all abnormal voltage coordinate points are polynomial fitted to obtain an abnormal voltage function.

[0227] Step S5, the heating time interval and the second stage heating time are obtained based on the normal voltage function, the abnormal voltage function, and the time threshold value obtaining method; step S5 includes the following sub-steps:

[0228] Step S501, the difference between all first screening thresholds and second screening thresholds is calculated, and is marked as a threshold difference;

[0229] Step S502, obtaining the mean value of the threshold difference absolute value is marked as the range difference threshold value;

[0230] Step S503, half of the range difference threshold value is marked as the resolution threshold value;

[0231] Step S504, in the voltage coordinate system, the difference between the ordinate value of the abnormal voltage function and the ordinate value of the normal voltage function at the same time is marked as the abnormal difference value;

[0232] Step S505, when the abnormal difference value is equal to the resolution difference value, the corresponding minimum value of the abscissa in the voltage coordinate system is marked as the heating time interval;

[0233] Step S506, after the heating time interval, the heating time of the resistance to the optimal temperature is marked as the second heating time;

[0234] Step S507, the second heating time is used as the experimental time to obtain the time threshold value by using the time threshold value obtaining method, and is marked as the second stage heating time.

[0235] Step S6, based on the first stage heating time, the heating time interval and the second stage heating time, the heating mode is set, which is marked as the low-power consumption heating mode; Step S6 includes the following sub-steps:

[0236] Step S601, set the gas detection sensor to start heating the resistance when working, stop heating after the first stage heating time, start heating after the heating time interval, and stop heating after the second stage heating time, then repeat the start heating after the heating time interval and the stop heating after the second stage heating time.

[0237] Step S7, the low-power consumption heating mode is used for gas detection; Step S7 includes the following sub-steps:

[0238] Step S701, in the low-power consumption heating mode and without containing the detection gas, the voltage on both sides of the resistance is marked as the detection voltage; the timing time started by the low-power consumption heating mode is marked as the working time;

[0239] Step S702, taking the working time as the X-axis data and the detection voltage as the Y-axis data, a plane rectangular coordinate system is established, which is marked as the detection coordinate system;

[0240] Step S703, taking the detection voltage and the corresponding working time as the ordinate and abscissa of the detection coordinate point respectively, all the detection coordinate points are plotted in the detection coordinate system;

[0241] Step S704, all the detection coordinate points are polynomially fitted to obtain the detection voltage function;

[0242] Step S705, obtaining the working time and the detection voltage of the gas detection sensor when working, and marking them as real-time time and real-time voltage respectively;

[0243] Step S706, substituting the real-time time into the detection voltage function to obtain the detection voltage marked as predicted normal voltage;

[0244] Step S707, calculating the difference between the real-time voltage and the predicted normal voltage, and marking it as detection difference;

[0245] Step S708, if the detection difference is greater than the resolution threshold, issuing a detection of the detection gas signal.

[0246] Those skilled in the art will understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media having computer-usable program code embodied in the medium. The storage media can be realized by any type of volatile or non-volatile storage devices or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk. These computer program instructions can also be stored in a computer readable storage medium which can guide the computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer readable storage medium produce the products including instruction devices, which realize the functions specified in the flowcharts Figure 1 one or more flows and / or blocks Figure 1 one or more flows and / or blocks

[0247] In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other manners. The embodiments described above are merely exemplary, for example, the division of the units is only a logical function division, and there can be another division manner in actual implementation; for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, or the among different units, can be indirect couplings or communication connections through some interfaces, communication interfaces, or a combination of the two, which can be electric, mechanical, or in other forms.

Claims

1. An ultra-low power consumption control method for gas detection sensors based on artificial intelligence, characterized in that, Includes the following steps: Obtain the optimal temperature for detecting resistance when the sensor is working, and mark it as the optimal temperature; The heating time for the first stage is obtained based on the optimal temperature and time threshold acquisition method. When there is no gas to detect, obtain the relationship function of the voltage across the resistor over time after heating to the optimal temperature and stopping, and mark it as the normal voltage function; When the detection gas is present, the relationship function of the voltage across the resistor versus time after heating to the optimal temperature and then stopping is obtained and marked as the abnormal voltage function. The heating time interval and the second-stage heating time are obtained based on the normal voltage function, abnormal voltage function and time threshold acquisition method. The heating method is set based on the first stage heating time, the heating time interval, and the second stage heating time, and is marked as a low-power heating method. Gas detection using a low-power heating method; The method for obtaining the optimal temperature and time thresholds to determine the first-stage heating time includes the following sub-steps: Obtain the heating time to reach the optimal temperature using resistance heating in the current environment, and mark it as the first heating time; The first heating time is used as the experimental time. The time threshold is obtained using the time threshold acquisition method and marked as the first stage heating time. Methods for obtaining time thresholds include: Obtain the first amount of experimental time; The experimental times were sorted in ascending order and labeled as Hs1 to Hs. i ; The smaller position ratio is obtained as: Ji = a1 * S1, where Ji is the smaller position ratio, a1 is the smaller coefficient, the range of a1 is (0, 0.5), and S1 is the first quantity; If the smaller position ratio is an integer, then Hs （Ji） The smaller position value is used as the identifier; if the smaller position ratio is not an integer, the integers to the left and right of Ji are obtained and labeled as Jiz and Jiy respectively, and Hs is calculated. （Jiz） With Hs （Jiy） The mean value is marked as the smaller position value; Determine if 0.5*S1 is an integer. If so, set Hs... （0.5*S1） Mark the value as the middle position; if not, obtain the integers on the left and right sides of 0.5*S1, and mark them as Jzz and Jzy respectively, and calculate Hs. （Jzz） With Hs （Jzz） The mean value is marked as the middle value. The larger coefficient is calculated as: a2 = [(1 / a1) - 1] * a1; where a2 is the larger coefficient. The larger position ratio is: Jx = a2*S1, where Jx is the smaller position ratio; If the larger position ratio is an integer, then Hs （Jx） The larger position value is marked; if the larger position ratio is not an integer, obtain the integers to the left and right of Jx, and mark them as Jxz and Jxy respectively, and calculate Hs. （Jxz） With Hs （Jxy） The mean value is marked as the smaller position value; The time threshold is calculated as: Sy = Wz + b1 * (Wx - Wi) / 2; where Sy is the time threshold, Wx is the larger position value, Wi is the smaller position value, Wz is the middle position value, and b1 is the threshold adjustment coefficient.

2. The ultra-low power consumption control method for an artificial intelligence-based gas detection sensor according to claim 1, characterized in that, In the absence of a detection gas, obtaining the relationship function of the voltage across the resistor over time after heating to the optimal temperature and stopping, and labeling it as the normal voltage function, includes the following sub-steps: The second number of times heating stops and timing begins at the optimal temperature is obtained, and the timing time at this time is marked as the first timing time. Every time interval T, the voltage across the resistor at this time is obtained and marked as the normal voltage. Mark the normal voltage at each identical interval as the normal interval voltage; The range of normal interval voltage is divided into c1 equal range intervals, which are marked as normal voltage division intervals; Count the frequency of each normal voltage division interval and mark it as the normal division frequency; A histogram is plotted with the normal interval voltage as the X-axis, the normal frequency division as the Y-axis, and the normal voltage division interval as the histogram interval. This histogram is then labeled as the normal voltage histogram. The second quantity is labeled S2; The first minimum frequency threshold is calculated as: F1 = d1 * S2 / c1; where F1 is the first frequency threshold and d1 is the proportion of the first minimum frequency. The normal division frequency that is less than or equal to the first smaller frequency threshold is marked as the first smaller frequency; Mark the leftmost and rightmost normal division frequencies of the normal power consumption histogram as normal side frequencies; Determine whether the normal side frequency is the first smallest frequency. If so, delete the portion of the normal voltage histogram where the normal side frequency is the first smallest frequency. Repeat the above operation on the normal voltage histogram after deletion until it is no longer the first smallest frequency. Mark the normal voltage histogram after stopping as the filtered normal voltage histogram. Obtain the maximum and minimum values ​​of the power consumption frequency abscissa in the normal voltage histogram and mark them as the first screening threshold and the second screening threshold, respectively.

3. The ultra-low power consumption control method for an artificial intelligence-based gas detection sensor according to claim 2, characterized in that, In the absence of a detection gas, obtaining the relationship function of the voltage across the resistor over time after heating to the optimal temperature and stopping, and labeling it as the normal voltage function, also includes the following sub-steps: The mean value of the normal interval voltage between the first screening threshold and the second screening threshold is calculated and marked as the proposed normal interval voltage; Obtain the proposed normal interval voltage for all normal interval voltages; Establish a Cartesian coordinate system with time as the X-axis data and voltage as the Y-axis data, and label it as the voltage coordinate system; Each proposed normal interval voltage and its corresponding first time are used as the ordinate and abscissa of the normal voltage coordinate point, respectively. All normal voltage coordinate points are plotted in the normal voltage coordinate system. The normal voltage function is obtained by polynomial fitting of all normal voltage coordinate points.

4. The ultra-low power consumption control method for an artificial intelligence-based gas detection sensor according to claim 3, characterized in that, When the detection gas is present, the relationship function of the voltage across the resistor versus time after heating to the optimal temperature and stopping is obtained and marked as the abnormal voltage function. This includes the following sub-steps: The third number of times heating stops and timing begins at the optimal temperature is obtained. This timing time is marked as the second timing time. At different second timing times, the detection gas is released, and the voltage across the resistor is obtained and marked as abnormal voltage. Mark the abnormal voltage at each second timing period as the abnormal interval voltage; The range of abnormal voltage intervals is divided into c2 equal intervals, which are marked as abnormal voltage intervals. The frequency of each abnormal voltage interval is counted and marked as the abnormal interval frequency; A histogram is drawn with the abnormal interval voltage as the X-axis, the abnormal frequency as the Y-axis, and the abnormal voltage interval as the histogram interval. This histogram is then marked as the abnormal voltage histogram. The third quantity is labeled S3; The second smallest frequency threshold is calculated as: F2 = d2 * S3 / c2; where F2 is the second frequency threshold and d2 is the proportion of the second smallest frequency. The frequency of anomalies that is less than or equal to the second smaller frequency threshold is marked as the second smaller frequency. The abnormal division frequencies of the leftmost and rightmost sides of the abnormal power consumption histogram are marked as abnormal side frequencies; Determine whether the frequency of the abnormal side is the second smallest frequency. If so, delete the portion of the abnormal voltage histogram where the frequency of the abnormal side is the second smallest frequency. Repeat the above operation on the abnormal voltage histogram after deletion until it is no longer the second smallest frequency. Mark the abnormal voltage histogram after stopping as the filtered abnormal voltage histogram. Obtain the maximum and minimum values ​​of the power consumption frequency abscissa of the abnormal voltage histogram and mark them as the third and fourth screening thresholds, respectively.

5. The ultra-low power consumption control method for an artificial intelligence-based gas detection sensor according to claim 4, characterized in that, When the detection gas is present, obtaining the relationship function of the voltage across the resistor over time after heating to the optimal temperature and stopping, and marking it as the abnormal voltage function, also includes the following sub-steps: The mean value of the abnormal interval voltage between the third screening threshold and the fourth screening threshold is calculated and marked as the proposed abnormal interval voltage. Obtain the proposed abnormal interval voltage for all abnormal interval voltages; Each proposed abnormal interval voltage and its corresponding second timing time are used as the vertical and horizontal coordinates of the abnormal voltage coordinate points, respectively, and all abnormal voltage coordinate points are plotted in the voltage coordinate system. The abnormal voltage function is obtained by performing polynomial fitting on all the abnormal voltage coordinate points.

6. The ultra-low power consumption control method for an artificial intelligence-based gas detection sensor according to claim 5, characterized in that, The heating time interval and the second-stage heating time are obtained based on the normal voltage function, abnormal voltage function, and time threshold acquisition method, including the following sub-steps: Calculate all the differences between the first and second screening thresholds and label them as threshold differences; The mean of the absolute values ​​of the threshold differences is marked as the range difference threshold. Mark half of the range difference threshold as the resolution threshold; In the voltage coordinate system, the difference between the ordinate value of the abnormal voltage function and the ordinate value of the normal voltage function at the same time is marked as the abnormal difference. When the abnormal difference is equal to the resolution difference, the minimum value of the corresponding horizontal axis in the voltage coordinate system is marked as the heating time interval. The heating time after the heating interval to heat the resistor to the optimal temperature is marked as the second heating time. The second heating time was used as the experimental time. The time threshold was obtained using the time threshold acquisition method and marked as the second stage heating time.

7. The ultra-low power consumption control method for an artificial intelligence-based gas detection sensor according to claim 6, characterized in that, The heating method is set based on the first-stage heating time, the heating time interval, and the second-stage heating time. The low-power heating method includes the following sub-steps: The gas detection sensor is set to start heating the resistor when it is working, stop heating after the first heating period, start heating again after the heating interval, stop heating again after the second heating period, and then repeat the cycle of starting heating after the heating interval and stopping heating after the second heating period.

8. The ultra-low power consumption control method for a gas detection sensor based on artificial intelligence according to claim 7, characterized in that, Gas detection using a low-power heating method includes the following steps: In the case of low-power heating mode and no detection gas, the voltage across the resistor is marked as the detection voltage; the timing time starting with low-power heating mode is marked as the working time. A Cartesian coordinate system is established with working time as the X-axis data and detected voltage as the Y-axis data, and this system is marked as the detection coordinate system. Using the detection voltage and the corresponding working time as the vertical and horizontal coordinates of the detection coordinate points respectively, all detection coordinate points are plotted in the detection coordinate system. The detection voltage function is obtained by performing polynomial fitting on all detection coordinate points; The working time and detection voltage of the gas detection sensor are obtained and marked as real-time time and real-time voltage, respectively. The detected voltage obtained by substituting the real-time time into the detected voltage function is marked as the predicted normal voltage; Calculate the difference between the real-time voltage and the predicted normal voltage, and mark it as the detection difference; If the detection difference is greater than the resolution threshold, a signal indicating that the gas has been detected is emitted.

9. An AI-based ultra-low power consumption control system for a gas detection sensor, used to implement the AI-based ultra-low power consumption control method for a gas detection sensor as described in any one of claims 1-8, characterized in that, It includes a temperature acquisition module, a time acquisition module, a first function acquisition module, a second function acquisition module, a heating interval acquisition module, a heating setting module, and a gas detection module; The temperature acquisition module is used to acquire the optimal detection temperature of the resistance when the sensor is working, and mark it as the optimal temperature; The time acquisition module is used to acquire the first stage heating time based on the optimal temperature and time threshold acquisition method. The first function acquisition module is used to acquire the relationship function of the voltage across the resistor over time after heating to the optimal temperature and stopping when there is no detection gas, and mark it as the normal voltage function; The second function acquisition module is used to acquire the relationship function of the voltage across the resistor over time after heating to the optimal temperature and stopping when the detection gas is present, and mark it as the abnormal voltage function; The heating interval acquisition module is used to acquire the heating time interval and the second stage heating time based on the normal voltage function, the abnormal voltage function and the time threshold acquisition method. The heating setting module is used to set the heating mode based on the first stage heating time, the heating time interval, and the second stage heating time, and to mark it as a low-power heating mode. The gas detection module is used to detect gas using a low-power heating method.

Images