Quality detection method for infant incubator

Abstract

The invention discloses an infant incubator quality detection method, which belongs to the technical field of detection, and comprises the steps of multi-channel temperature and humidity acquisition, self-inspection and self-calibration, reference channel selection, weight distribution, abnormal supplementary acquisition and consistency judgment, volatility evaluation and data reporting. According to the invention, high-precision, multi-angle, dynamic and credible environment quality detection of the infant incubator can be realized, the reliability of deviation judgment is remarkably improved, the overall precision is prevented from being lowered by a certain abnormal channel, a wrong reference value is prevented from being used for calculation, and the accuracy of the infant incubator is improved. And the overall fault tolerance and robustness of the system are improved.

Description

Technical Field

[0001] The present invention belongs to the field of detection technology, and in particular relates to a method for detecting the quality of an infant incubator. Background Art

[0002] Infant incubators are widely used in the field of neonatal intensive care. Their core indicators include precise control and stability of temperature and humidity.

[0003] Existing incubator quality testing methods typically rely on single-point temperature and humidity sensors combined with timed readings and simple average calculations, which presents the following issues:

[0004] Due to the lack of a multi-channel data consistency analysis mechanism, errors cannot be effectively identified and compensated.

[0005] The sampling data was not screened for stability during the testing cycle, resulting in occasional deviations affecting the final evaluation results.

[0006] A means of analyzing temperature and humidity fluctuation trends has not been established, making it difficult to identify environmental anomalies or insufficient stability in a short period of time. Summary of the Invention

[0007] The purpose of the present invention is to provide a method for quality inspection of infant incubators, which solves the technical problem of achieving consistency evaluation and stability determination of multi-channel temperature data by introducing dynamic correction of relative deviations between sensors and a weighted fusion strategy.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] A method for testing the quality of an infant incubator comprises the following steps:

[0010] Step 1: After the detector is powered on, it is initialized and self-tests the temperature and humidity sensors and the communication channel respectively. If an abnormality occurs, the self-test process is interrupted and a self-test report is generated.

[0011] Step 2: The detector determines whether to enter the self-calibration process based on the consistency of the initial readings of each temperature sensor. If not, it performs five simultaneous samplings, calculates the maximum temperature difference, generates and writes the compensation matrix; if the correction exceeds the limit, the calibration is marked as failed;

[0012] Step 3: The detector sets the RTC timer to trigger the detection cycle at the preset interval;

[0013] Step 4: Before the test cycle, the detector samples each temperature channel multiple times, selects the channel with the smallest standard deviation as a reference, calculates the deviation of other temperature channels, and dynamically assigns trust weights based on the deviation results; each temperature channel corresponds to a temperature sensor;

[0014] Step 5: The detector performs multiple temperature and humidity sampling within the detection cycle, performs supplementary sampling and invalidation processing on abnormal sampling values, and reduces the weight of the invalid temperature channel;

[0015] Step 6: The detector calculates the average of the effective temperature values, obtains the temperature representative value and the humidity representative value, and determines the eligibility of the temperature consistency;

[0016] Step 7: The detector writes the temperature representative value and the humidity representative value into the circular buffer, calculates the temperature fluctuation range and the humidity fluctuation range, and makes a stability determination based on the calculation results;

[0017] Step 8: The detector locally caches all types of data generated in each round of detection cycle, and uploads all types of data to the host computer after the communication connection with the host computer is established;

[0018] Step 9: The host computer generates a test report in PDF / CSV format and displays the historical quality test results in charts.

[0019] Preferably, when executing step 1, specifically:

[0020] After the detector is powered on, it is initialized and self-tests the five temperature sensors and humidity sensors respectively. This includes obtaining the status register and initial readings of each sensor channel and checking whether the communication channel with the host computer is normal.

[0021] If the self-test fails, an alarm message is generated, the self-test process is interrupted, and a self-test report is generated.

[0022] Preferably, when executing step 2, specifically:

[0023] The detector determines whether to enter the power-on self-calibration mode based on the consistency between the initial readings of the five temperature sensors. If not, proceed to step 3. If yes, enter the power-on self-calibration mode, which specifically includes:

[0024] At the same time, the five temperature sensors are triggered to sample five times continuously to construct the original temperature data set {T a1 ...T a5}, calculate the maximum temperature difference ΔT of the original temperature data set a max: If the maximum temperature difference ΔT a If max is less than the preset threshold, it will be recorded as "preliminary verification passed";

[0025] On the contrary, according to the maximum temperature difference ΔT a max and the deviation of each channel temperature sensor to generate the offset compensation matrix and write it into the compensation register of each temperature sensor channel;

[0026] If the number of deviation corrections for a temperature sensor channel exceeds the correction upper limit, a "calibration failed" message is generated.

[0027] Preferably, when executing step 4, specifically:

[0028] During the interval of each detection cycle, the detector performs high-speed sampling on each of the five temperature sensors five times to obtain the periodic calibration data set {t i1 ...t i5};

[0029] Calculate the standard deviation {σ1…σ1} of each temperature sensor, select the minimum value σmin, and use the temperature sensor corresponding to the minimum value σmin as the reference sensor T for this round of detection cycle ref ;

[0030] Calculate the temperature of the other 4 sensors relative to the reference sensor T ref Deviation Δt j =t j –t ref , write the result into the sensor trust table;

[0031] According to the historical stability score and the current deviation of each temperature sensor, the temperature sensor is assigned a weight W in the sensor trust table. i , set W ref The highest weight.

[0032] Preferably, when executing step 5, specifically:

[0033] At the beginning of the detection cycle, the detector samples each sensor three times in turn, and obtains the temperature set {T i1 ,T i2 ,T i3} and humidity set {H1,H2,H3}, calculate the standard deviation of temperature set and humidity set respectively, denoted as σ T and σ H ;

[0034] If the deviation of a sensor exceeds the threshold, re-sampling is performed. If it still exceeds the threshold after re-sampling, the sensor is marked as "failed in this round" and the weight of the sensor in the sensor trust table is lowered.

[0035] Preferably, when executing step 6, specifically:

[0036] After completing a test cycle, the detector will collect the temperature data {T i1 ,T i2 ,T i3} to calculate the average value of the valid sampling values ​​and obtain the mean value set {Tb1 ...T b5};

[0037] Calculate the consistency mean Tavg=(ΣT bi ) / 5, calculate the maximum deviation ΔT based on the consistency mean Tavg b max=max|T bi –Tavg|;

[0038] If ΔT b If max exceeds the preset threshold, it will record "this round of detection cycle consistency failed" and output an alarm message; otherwise, it will be marked as "this round of detection cycle consistency passed";

[0039] Calculate the representative temperature T final =Σ(W·T);

[0040] Calculate the mean of the humidity set {H1, H2, H3} to obtain the humidity mean H, and use the mean H as the representative humidity, denoted as H final .

[0041] Preferably, when executing step 7, specifically:

[0042] The detector sets the T of this test cycle final and H final , respectively appended to two 30-capacity circular buffers, and calculate the temperature fluctuation value ΔT respectively window and humidity band value ΔH window :

[0043] ΔT window =max(T final30 )–min(T final30 );

[0044] ΔH window =max(H final30 )–min(H final30 );

[0045] Judgment: When the temperature fluctuation value ΔT window When the temperature fluctuation exceeds the preset threshold, it will be marked as "this round of temperature stability is unqualified";

[0046] When the humidity fluctuation value ΔH window When the humidity fluctuation threshold is exceeded, it will be marked as “this round of humidity stability is unqualified”.

[0047] The present invention addresses the technical problem of achieving consistency assessment and stability determination for multi-channel temperature data by introducing a dynamic correction of relative deviations between sensors and a weighted fusion strategy. Through multi-channel temperature and humidity sensor acquisition, standard deviation-based reference channel screening, dynamic deviation correction, weighted stability assessment, and fluctuation trend determination, the present invention enables high-precision, multi-angle, and dynamically reliable environmental quality testing for infant incubators. A minimum standard deviation screening mechanism automatically selects the most stable and least noisy channel as the reference channel, significantly improving the reliability of deviation determination and preventing overall accuracy from being degraded by an abnormal channel. In the event of sudden changes in the external environment or a temporary failure of a channel, a stable channel can be immediately reselected, preventing incorrect reference values ​​from being used in calculations and improving the overall fault tolerance and robustness of the system. The present invention adjusts the trust weights of other channels based on their deviations from the reference channel and historical scores, constructing a dynamic trust model that makes the fusion results more statistically significant and enhances the long-term stability of the system. This eliminates the problem of "false consistency" judgments when multiple data channels deviate but are close to each other. The judgment process is based on volatility rather than the value itself, avoiding the false consistency problem of multiple abnormal values ​​"covering each other." BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is the main flow chart of the present invention;

[0049] Figure 2 is a flow chart of step 2 of the present invention;

[0050] Figure 3 is a flow chart of step 4 of the present invention;

[0051] Figure 4 is a flow chart of steps 6 and 7 of the present invention;

[0052] Figure 5 It is a schematic block diagram of the detector in this embodiment. DETAILED DESCRIPTION

[0053] Depend on Figure 1-Figure 5 A method for testing the quality of an infant incubator is shown, comprising the following steps:

[0054] Step 1: After the detector is powered on, it is initialized and self-tests the temperature and humidity sensors and the communication channel respectively. If an abnormality occurs, the self-test process is interrupted and a self-test report is generated.

[0055] When executing step 1, specifically:

[0056] After the detector is powered on, it is initialized and self-tests the five temperature sensors and humidity sensors respectively. This includes obtaining the status register and initial readings of each sensor channel and checking whether the communication channel with the host computer is normal.

[0057] If the self-test fails, an alarm message is generated, the self-test process is interrupted, and a self-test report is generated.

[0058] In this embodiment, after the detector is powered on, its main control board will read the status registers and initial measurement values ​​of 5 PT100 temperature sensors (mainly PT100), 1 SHT35 humidity sensor and UART / BLE communication link in sequence.

[0059] In this embodiment, the main control board includes a main control chip, a communication chip, a power module and a memory. The main control chip adopts an STM32F103C8T6 type single-chip microcomputer, the communication chip adopts a CH340X communication chip, the power module adopts a TP4056 type single-cell lithium-ion battery constant current / constant voltage linear charger, and the memory adopts a 24C02 type EEPROM storage chip.

[0060] The main control chip uses I 2 The C / SPI interface obtains the internal status flag of each sensor and the first set of temperature and humidity readings, and then verifies the link connectivity with the host computer through the communication unit self-test handshake packet.

[0061] The self-test results (pass / fail) are written to the local EEPRAM, and a self-test report containing a timestamp and status code is generated in the storage unit.

[0062] Step 2: The detector determines whether to enter the self-calibration process based on the consistency of the initial readings of each temperature sensor. If not, it performs five simultaneous samplings, calculates the maximum temperature difference, generates and writes the compensation matrix; if the correction exceeds the limit, the calibration is marked as failed;

[0063] When entering the self-calibration process, the user needs to place the detector in a constant temperature box for calibration to ensure the calibration quality.

[0064] When executing step 2, specifically:

[0065] The detector determines whether to enter the power-on self-calibration mode based on the consistency between the initial readings of the five temperature sensors. If not, proceed to step 3. If yes, enter the power-on self-calibration mode, which specifically includes:

[0066] At the same time, the five temperature sensors are triggered to sample five times continuously to construct the original temperature data set {T a1 ...T a5}, calculate the maximum temperature difference ΔT of the original temperature data set a max: If the maximum temperature difference ΔT a If max is less than the preset threshold, it will be recorded as "preliminary verification passed";

[0067] On the contrary, according to the maximum temperature difference ΔT amax and the deviation of each channel temperature sensor to generate the offset compensation matrix and write it into the compensation register of each temperature sensor channel;

[0068] If the number of deviation corrections for a temperature sensor channel exceeds the correction upper limit, a "calibration failed" message is generated.

[0069] After the detector is turned on, it will first compare the initial readings of the five temperature sensors for consistency; if the maximum temperature difference exceeds the threshold, it will trigger the five sensors to collect the temperature 5 times at the same time, forming {T a1 …T a5} set and calculate ΔT a max, and use this to generate the offset compensation matrix and write it back to the sensor compensation register.

[0070] This embodiment uses multiple redundant sampling and the range method to evaluate the systematic error of each probe and eliminates the zero drift between channels through online offset correction. The offset matrix and the correction count are written into the firmware to cumulatively monitor the sensor performance.

[0071] like:

[0072] There are 5 temperature sensors on the device: T1, T2, T3, T4, and T5. The initial temperatures read after power on are: T1 = 36.5, T2 = 36.7, T3 = 36.4, T4 = 36.9, and T5 = 36.6.

[0073] In this embodiment, the maximum value 36.9 (T4) is first found, and then the minimum value 36.4 (T3) is found to calculate the maximum temperature difference ΔT a max=36.9–36.4=0.5℃。

[0074] The threshold is set to 0.3℃, 0.5℃>0.3℃, so the self-calibration process will be entered after power on. If the calculated result is less than 0.3℃, it will not enter.

[0075] The default reference value Tref at startup is the average value Tavg of all temperature values:

[0076] Tavg(36.5+36.7+36.4+36.9+36.6) / 5=36.62℃.

[0077] The output of all channels after calibration should be close to 36.62℃. Calculate the offset value of each channel, which is Tavg – original temperature:

[0078] T1 is +0.12, T2 is -0.08, T3 is +0.22, T4 is -0.28, and T5 is +0.02.

[0079] The offset compensation matrix is ​​as follows:

[0080] Ofset_Matrix=[+0.12,-0.08,+0.22,-0.28,+0.02].

[0081] Each sensor channel automatically adds the corresponding offset to its original sampling value during subsequent output to obtain a corrected value. For example, if the sampling temperature of T1 is 36.8°C, the output is 36.8+0.12=36.92°C.

[0082] Step 3: The detector sets the RTC timer to trigger the detection cycle at the preset interval;

[0083] The RTC timer is a built-in RTC timer in the main control unit, which automatically triggers a complete detection cycle every preset time (such as 1 minute).

[0084] Step 4: Before the test cycle, the detector samples each temperature channel multiple times, selects the channel with the smallest standard deviation as a reference, calculates the deviation of other temperature channels, and dynamically assigns trust weights based on the deviation results; each temperature channel corresponds to a temperature sensor;

[0085] When executing step 4, specifically:

[0086] During the interval of each detection cycle, the detector performs high-speed sampling on each of the five temperature sensors five times to obtain the periodic calibration data set {t i1 ...t i5};

[0087] Calculate the standard deviation {σ1…σ1} of each temperature sensor, select the minimum value σmin, and use the temperature sensor corresponding to the minimum value σmin as the reference sensor T for this round of detection cycle ref ;

[0088] Calculate the temperature of the other 4 sensors relative to the reference sensor T ref Deviation Δt j =t j –t ref , write the result into the sensor trust table;

[0089] According to the historical stability score and the current deviation of each temperature sensor, the temperature sensor is assigned a weight W in the sensor trust table. i , set W ref The highest weight.

[0090] Before the detection cycle, perform 5 high-speed sampling on each temperature channel and calculate the respective standard deviation σ i , select the channel where σmin is located as the reference T ref ; Calculate other channels and T ref The average deviation Δt j, and update the trust weight W based on the historical stability score i (W ref Highest).

[0091] The most stable channel is identified based on the minimum variance principle, and the Δtj value is written into the trust table to quantify the real-time accuracy of each sensor. The trust weight is used in the subsequent fusion algorithm to reduce the impact of abnormal fluctuating channels.

[0092] Step 5: The detector performs multiple temperature and humidity sampling within the detection cycle, performs supplementary sampling and invalidation processing on abnormal sampling values, and reduces the weight of the invalid temperature channel;

[0093] When executing step 5, specifically:

[0094] At the beginning of the detection cycle, the detector samples each sensor three times in turn, and obtains the temperature set {T i1 ,T i2 ,T i3} and humidity set {H1,H2,H3}, calculate the standard deviation of temperature set and humidity set respectively, denoted as σ T and σ H ;

[0095] If the deviation of a sensor exceeds the threshold, re-sampling is performed. If it still exceeds the threshold after re-sampling, the sensor is marked as "failed in this round" and the weight of the sensor in the sensor trust table is lowered.

[0096] In this embodiment, each temperature and humidity sensor is sampled 3 times during the detection period to form a temperature {T i1 ,T i2 ,T i3} and humidity {H1,H2,H3} data sets, calculate the standard deviation σ T / σ H If any reading deviation exceeds the threshold, a single-point supplementary sampling is triggered. If the supplementary sampling is still abnormal, the channel is marked as invalid for this round and its weight is lowered.

[0097] Transient noise is eliminated through multiple redundant sampling combined with σ-threshold outlier detection. After the resampled data flows back, σ is calculated again to ensure the reliability of the current round of data, and the channel health is updated in real time in the trust table, which improves the measurement robustness, effectively eliminates occasional errors, and reduces the impact of a single failure on the overall evaluation.

[0098] Step 6: The detector calculates the average of the effective temperature values, obtains the temperature representative value and the humidity representative value, and determines the eligibility of the temperature consistency;

[0099] When executing step 6, specifically:

[0100] After completing a test cycle, the detector will collect the temperature data {T i1 ,T i2 ,T i3} to calculate the average value of the valid sampling values ​​and obtain the mean value set {T b1 ...T b5};Calculate the consistency mean Tavg=(ΣT bi ) / 5, calculate the maximum deviation ΔT based on the consistency mean Tavg b max=max|T bi –Tavg|;

[0101] If ΔT b If max exceeds the preset threshold, it will record "this round of detection cycle consistency failed" and output an alarm message; otherwise, it will be marked as "this round of detection cycle consistency passed";

[0102] Calculate the representative temperature T final =Σ(W·T);

[0103] In this embodiment, in multiple temperature channels, according to the trustworthiness of each sensor (weight W i ) its temperature measurement value T i Weighted, a reliable result T representing the overall temperature is obtained final , assigning higher weight W to sensors with high stability and small error i , the weight of channels with poor stability, frequent supplementary sampling, and abnormalities is lower, which can reduce the influence of unreliable channels on the representative value T final interference.

[0104] Calculate the mean of the humidity set {H1, H2, H3} to obtain the humidity mean H, and use the mean H as the representative humidity, denoted as H final .

[0105] In this embodiment, the effective temperature samples are averaged to obtain T b1 …T b1 , calculate the overall mean Tavg and maximum deviation ΔT b max; if ΔT b If max exceeds the limit, an alarm will be issued, otherwise the trust weight W i Calculate the weighted representative temperature T final ; At the same time, the average value of the humidity sample is H final , first make the extreme difference consistency judgment, then fuse each channel based on weighted average, Tavg / ΔT b max and T final At the same time, the cache and output queue are written for subsequent links to ensure that the multi-channel temperatures are consistent within the threshold range before fusion, ensuring that the final output meets both stability and the advantages of multi-point measurement.

[0106] Step 7: The detector writes the temperature representative value and the humidity representative value into the circular buffer, calculates the temperature fluctuation range and the humidity fluctuation range, and makes a stability determination based on the calculation results;

[0107] When executing step 7, specifically:

[0108] The detector sets the T of this test cycle final and H final , respectively appended to two 30-capacity circular buffers, and calculate the temperature fluctuation value ΔT respectively window and humidity band value ΔH window :

[0109] ΔT window =max(T final30 )–min(T final30 );

[0110] ΔH window =max(H final30 )–min(H final30 );

[0111] Judgment: When the temperature fluctuation value ΔT window When the temperature fluctuation exceeds the preset threshold, it will be marked as "this round of temperature stability is unqualified";

[0112] When the humidity fluctuation value ΔH window When the humidity fluctuation threshold is exceeded, it will be marked as “this round of humidity stability is unqualified”.

[0113] This embodiment adopts the time series sliding window statistics method, uses the latest 30 data for fluctuation amplitude assessment, writes the results into the stability flag, monitors environmental stability from the perspective of overall trend, avoids misjudgment caused by single-point anomalies, and supports dual observation of short-term fluctuations and long-term trends.

[0114] Step 8: The detector locally caches all types of data generated in each round of detection cycle, and uploads all types of data to the host computer after the communication connection with the host computer is established;

[0115] Various data including temperature representative value T final , Humidity representative value H final , consistency judgment results, temperature fluctuation value ΔT window , humidity fluctuation value ΔH window , which sensors are judged to be failed in this round of detection, the current weight W of each channel i , as well as the reference channel mark Tref, self-test failure information, calibration failure information, alarm for unqualified consistency / stability in this round, time information of each record, original temperature set {T i1,T i2 ,T i3}, original humidity set {H1,H2,H3}, periodic calibration data set {t i1 ...t i5}.

[0116] Step 9: The host computer generates a test report in PDF / CSV format and displays the historical quality test results in charts.

[0117] The host computer generates a quality inspection report in PDF / CSV format based on the historical data set and displays it in the form of icons.

[0118] The present invention addresses the technical problem of achieving consistency assessment and stability determination for multi-channel temperature data by introducing a dynamic correction of relative deviations between sensors and a weighted fusion strategy. Through multi-channel temperature and humidity sensor acquisition, standard deviation-based reference channel screening, dynamic deviation correction, weighted stability assessment, and fluctuation trend determination, the present invention enables high-precision, multi-angle, and dynamically reliable environmental quality testing for infant incubators. A minimum standard deviation screening mechanism automatically selects the most stable and least noisy channel as the reference channel, significantly improving the reliability of deviation determination and preventing overall accuracy from being degraded by an abnormal channel. In the event of sudden changes in the external environment or a temporary failure of a channel, a stable channel can be immediately reselected, preventing incorrect reference values ​​from being used in calculations and improving the overall fault tolerance and robustness of the system. The present invention adjusts the trust weights of other channels based on their deviations from the reference channel and historical scores, constructing a dynamic trust model that makes the fusion results more statistically significant and enhances the long-term stability of the system. This eliminates the problem of "false consistency" judgments when multiple data channels deviate but are close to each other. The judgment process is based on volatility rather than the value itself, avoiding the false consistency problem of multiple abnormal values ​​"covering each other."

Claims

1. A method for testing the quality of an infant incubator, characterized by: The steps include: Step 1: After the detector is powered on, it is initialized and self-tests the temperature and humidity sensors and the communication channel respectively. If an abnormality occurs, the self-test process is interrupted and a self-test report is generated. Step 2: The detector determines whether to enter the self-calibration process based on the consistency of the initial readings of each temperature sensor. If not, it performs five simultaneous samplings, calculates the maximum temperature difference, generates and writes the compensation matrix; if the correction exceeds the limit, the calibration is marked as failed; Step 3: Detector Settings A timer that triggers a detection cycle at preset intervals; Step 4: The detector samples each temperature channel multiple times before the test cycle, selects the channel with the smallest standard deviation as the reference, calculates the deviation of other temperature channels, and dynamically assigns trust weights based on the deviation results; Each temperature channel corresponds to a temperature sensor; Step 5: The detector performs multiple temperature and humidity sampling within the detection cycle, performs supplementary sampling and invalidation processing on abnormal sampling values, and reduces the weight of the invalid temperature channel; Step 6: The detector calculates the average of the effective temperature values, obtains the temperature representative value and the humidity representative value, and determines the eligibility of the temperature consistency; Step 7: The detector writes the temperature representative value and the humidity representative value into the circular buffer, calculates the temperature fluctuation range and the humidity fluctuation range, and makes a stability determination based on the calculation results; Step 8: The detector locally caches all types of data generated in each round of detection cycle, and uploads all types of data to the host computer after the communication connection with the host computer is established; Step 9: Generate host computer Format inspection report and display historical quality inspection results in chart form.

2. The method for testing the quality of an infant incubator according to claim 1, wherein: When executing step 1, specifically: After the detector is powered on, it is initialized and self-tests the five temperature sensors and humidity sensors respectively. This includes obtaining the status register and initial readings of each sensor channel and checking whether the communication channel with the host computer is normal. If the self-test fails, an alarm message is generated, the self-test process is interrupted, and a self-test report is generated.

3. The method for testing the quality of an infant incubator according to claim 2, wherein: When executing step 2, specifically: The detector determines whether to enter the power-on self-calibration mode based on the consistency between the initial readings of the five temperature sensors. If not, proceed to step 3. If yes, it will enter the power-on self-calibration mode, including: Simultaneously trigger 5 temperature sensors to sample 5 times continuously to build the original temperature data set Calculate the maximum temperature difference of the original temperature data set If the maximum temperature difference If the value is less than the preset threshold, it will be recorded as "preliminary calibration passed"; On the contrary, according to the maximum temperature difference The deviation of each channel temperature sensor is used to generate an offset compensation matrix and write it into the compensation register of each temperature sensor channel; If the number of deviation corrections for a temperature sensor channel exceeds the correction upper limit, a "calibration failed" message is generated.

4. The method for testing the quality of an infant incubator according to claim 3, wherein: When executing step 4, specifically: During the interval of each detection cycle, the detector performs high-speed sampling on each of the five temperature sensors five times to obtain a periodic calibration data set. Calculate the standard deviation of each temperature sensor Select the minimum value Minimum The corresponding temperature sensor is used as the reference sensor for this round of detection cycle Calculate the temperature of the other 4 sensors relative to the reference sensor Deviation Write the result to the sensor trust table; Assign weights to temperature sensors in the sensor trust table based on historical stability scores and current deviations of each temperature sensor. set up The highest weight.

5. The method for testing the quality of an infant incubator according to claim 4, wherein: When executing step 5, specifically: At the beginning of the detection cycle, the detector samples each sensor three times in turn to obtain the temperature set of the detection cycle. and humidity collection Calculate the standard deviation of the temperature set and humidity set respectively, and record them as and If the deviation of a sensor exceeds the threshold, re-sampling is performed. If the deviation still exceeds the threshold after re-sampling, the sensor is marked as "failed in this round" and its weight in the sensor trust table is lowered.

6. The method for testing the quality of an infant incubator according to claim 5, wherein: When executing step 6, specifically: After completing a test cycle, the detector collects the temperature The effective sampling values ​​in are averaged to obtain the mean value set Calculate the mean consistency According to the mean of consistency Calculate the maximum deviation like If the preset threshold is exceeded, "Consistency of this round of detection cycle failed" will be recorded and an alarm message will be output; Otherwise, it will be marked as "this round of test cycle consistency passed"; Calculate representative temperature Humidity Collection Perform mean calculation to obtain the mean humidity The mean As the representative humidity, 7. The method for testing the quality of an infant incubator according to claim 6, wherein: When executing step 7, specifically: The detector will and Append to two 30-entry circular buffers and calculate the temperature fluctuation values ​​respectively and humidity band values Judgment: When the temperature fluctuation value When the temperature fluctuation exceeds the preset threshold, it will be marked as "this round of temperature stability is unqualified"; When humidity fluctuates When the humidity fluctuation threshold is exceeded, it will be marked as "this round of humidity stability is unqualified".

Images