A management method of an air conditioner of an internet of things

Abstract

The application relates to the field of air conditioner management, and discloses a management method of an air conditioner of an Internet of Things, which performs environment monitoring through a multi-modal sensor arranged in a high-dust area; the multi-modal sensor comprises a temperature sensor, a humidity sensor and an infrared thermal imager; and the dust concentration is monitored in real time; when the dust concentration exceeds a set dust concentration threshold value, compressed air blowing or ultrasonic vibration cleaning is triggered; after cleaning, sensor response testing is performed; if the actual time consumption of temperature reaching a set falling difference value exceeds set time consumption, it is determined that the sensor is abnormal, a standby sensor is started, and an alarm is given; finally, the temperature sensor, the infrared thermal imager and air outlet temperature data of the air conditioner are fused according to respective weight coefficients, and data of the humidity sensor is combined, so that a corrected environment temperature value is output; if a single sensor reading deviates from the fusion result by more than a preset temperature range interval, the standby sensor is started, and an alarm is given, so that accurate control of the environment temperature is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of air conditioner management, in particular to a management method of an air conditioner of Internet of Things. BACKGROUND

[0002] In recent years, with the rapid development of Internet of Things technology, intelligent air conditioner management systems have been widely used in industrial, commercial and home scenarios. Such systems achieve remote monitoring and automatic adjustment of air conditioning equipment by deploying temperature and humidity sensors, network communication modules and intelligent control algorithms, in order to achieve energy saving and improve comfort. However, in high dust environments such as welding workshops, textile factories and mines, existing systems still face serious technical defects, mainly manifested in that dust adhesion leads to abnormal sensor readings, and then causes air conditioner control failure. The specific analysis process is as follows:

[0003] Metal dust, fiber dust and other industrial environment in the surface of temperature and humidity sensor, form heat insulation layer or hygroscopic layer, cause temperature reading high, humidity detection distortion. For example, dust coverage makes temperature sensor response delay, system misjudges that environment temperature is not up to standard, continuously overcools, causes energy waste.

[0004] And although some systems use IP67 dustproof sensors and regular air blowing cleaning, but still have the following problems:

[0005] IP67 protection limitation: only prevent dust from entering the inside of the sensor, but cannot avoid the interference of dust accumulation on the surface of the shell on heat conduction.

[0006] Cleaning is not thorough: high viscosity dust such as oily metal chips is difficult to remove by air blowing, and when the cleaning cycle is unreasonable, the sensor still continues to misreport between two cleanings.

[0007] Improper installation position: the sensor is deployed in the airflow dead angle or vibration area, which accelerates dust adhesion and is difficult to clean. SUMMARY

[0008] The purpose of the present application is to provide a management method of an air conditioner of Internet of Things, which solves at least one of the above technical problems.

[0009] The purpose of the present application can be achieved by the following technical solutions:

[0010] A management method of an air conditioner of Internet of Things, comprising the following steps:

[0011] Step one, environment monitoring by multi-modal sensor deployed in high dust area; including temperature sensor, humidity sensor and infrared thermal imager;

[0012] Step two, real-time monitoring of dust concentration, when the dust concentration exceeds the set dust concentration threshold, trigger compressed air blowing or ultrasonic vibration cleaning;

[0013] After cleaning, perform a response test of the temperature sensor, if the actual time taken for the temperature to reach the set drop difference exceeds the set time, determine that the temperature sensor is abnormal, start the backup temperature sensor and alarm at the same time;

[0014] Step three, fuse the temperature sensor, infrared thermal imager and air outlet temperature data according to the respective weight coefficients, and combine the humidity sensor data to output the corrected ambient temperature value;

[0015] If the difference between the temperature sensor reading and the corrected ambient temperature value exceeds the preset temperature range interval, start the backup temperature sensor and alarm.

[0016] Further, the step two further includes:

[0017] A mobile cleaning robot is used to regularly patrol and clean the surface of each temperature sensor along the preset path;

[0018] After cleaning, compare the change rate of the temperature sensor data before and after cleaning, if it is lower than the set change rate threshold, determine that the cleaning operation is invalid, and notify manual intervention.

[0019] Further, the method further includes:

[0020] Through a pre-trained long short-term memory (LSTM) neural network model, predict the influence trend of dust accumulation on the temperature sensor, and dynamically adjust the air conditioning refrigeration power;

[0021] In the area where the dust concentration continues to exceed the standard, increase the weight of the air outlet temperature of the air conditioner, and reduce the weight of the temperature sensor.

[0022] Further, the process of fusing the temperature sensor, infrared thermal imager and air outlet temperature data of the air conditioner according to the respective weight coefficients and combining the humidity sensor data to output the corrected ambient temperature value is:

[0023] After preprocessing the data of the temperature sensor, infrared thermal imager, air outlet temperature of the air conditioner and humidity sensor, dimensionless first temperature value , second temperature value , third temperature value , humidity value S are obtained respectively;

[0024] Substitute into the formula:

[0025] The corrected ambient temperature value is calculated.

[0026] In the formula, ; , respectively represent the maximum value, the minimum value of the first, second, third temperature value at the first sampling time; , , , is a weight coefficient, determined based on historical data analysis; represents the total number of samplings within a set period of time from the current time, represents the first, , is a judgment function about the humidity value, is an adjustment parameter greater than 0, determined based on historical data analysis.

[0027] Further, the expression of the judgment function about the humidity value is as follows:

[0028] ;

[0029] In the formula, represents the lower limit of the measured humidity value, , represents the pre-set humidity interval limit, , , represents the proportional coefficient corresponding to different humidity intervals, obtained by fitting according to experiments or historical data.

[0030] Further, the process of predicting the influence trend of dust accumulation on the temperature sensor through the pre-trained long short-term memory (LSTM) neural network model is as follows:

[0031] The first, second, third temperature values, humidity value and dust concentration are input into the pre-trained long short-term memory (LSTM) neural network model, and the accumulated dust amount of the temperature sensor per unit time, i.e. the dust accumulation rate G, is output.

[0032] The dust accumulation rate G is compared with the pre-set dust accumulation threshold G0. If G≥G0, it is predicted that the influence trend of dust accumulation on the temperature sensor is abnormal.

[0033] Otherwise, it is predicted that the influence trend of dust accumulation on the temperature sensor is normal.

[0034] Further, the process of dynamically adjusting the air conditioning refrigeration power is as follows:

[0035] ​If G is greater than or equal to G0 and the difference between the temperature sensor reading and the corrected ambient temperature value in a historical time period is in an increasing state, the cooling power of the air conditioner is reduced; otherwise, the current cooling power of the air conditioner is maintained.

[0036] Further, the process of judging that the dust concentration continues to exceed the standard is:

[0037] The average dust concentration in a historical period in the current area is compared with the preset warning value;

[0038] If the average dust concentration is greater than or equal to the preset warning value, and the dust accumulation rate G of at least one temperature sensor in the current area is greater than or equal to G0, it is judged that the current dust concentration continues to exceed the standard, and the current area is marked as a dust concentration continues to exceed the standard area.

[0039] The beneficial effects of the present application are:

[0040] (1) Through multi-modal sensor cooperative monitoring, combined with complex correction calculation method, the influence of various factors on temperature measurement in high dust environment is comprehensively considered, in the calculation of corrected temperature value, the temperature, humidity data, the data dispersion degree and the weight of different sensors are considered, the error caused by the dust interference of the sensor is effectively reduced, the monitored environmental temperature is closer to the true value, the accurate temperature data is provided for the air conditioner, the cooling power can be accurately adjusted, the excessive cooling caused by temperature misjudgment is avoided, the accuracy of temperature control is improved, and the demand of users for comfort is met.

[0041] (2) A variety of cleaning methods and strict sensor state monitoring mechanism are adopted, the normal operation of the sensor in high dust environment is ensured; when the dust concentration exceeds the standard, compressed air blowing, ultrasonic vibration cleaning are triggered, the mobile cleaning robot can also regularly patrol and clean, the influence of dust on the sensor is reduced; after cleaning, the sensor response test is carried out, once the abnormality is found, the standby sensor is started and an alarm is given, and the dust accumulation influence trend is predicted combining with the historical data, the cooling power of the air conditioner is dynamically adjusted, so that the stability and reliability of the sensor data are ensured, even in severe high dust environment, the system failure risk is reduced, and the service life of the equipment is prolonged.

[0042] (3) Based on the accurate environmental temperature data, the air conditioner can adjust the operation mode according to the actual demand, avoid unnecessary energy consumption; in the area where the dust concentration continues to exceed the standard, the control mode is switched and the weight is adjusted, the air conditioner operation strategy is optimized, and the energy waste caused by sensor error is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0043] The present application will be further described below with reference to the accompanying drawings.

[0044] Figure 1A flow chart of the method of the present application. DETAILED DESCRIPTION

[0045] 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 work fall within the protection scope of the present application.

[0046] Please refer to Figure 1 The present application is a management method of an air conditioner in an Internet of Things, comprising the following steps:

[0047] Step one, environmental monitoring by multi-modal sensors deployed in high-dust areas; including temperature sensors, humidity sensors and infrared thermal imagers; including: open PT100 temperature sensors, coated with a metal dust-repellent coating, directly exposed to the airflow to reduce heat conduction delay; capacitive humidity sensors, integrated with electrostatic dust removal units, periodically releasing high-voltage pulses to clean the electrode surface dust; infrared thermal imagers for cross-verification of the authenticity of temperature sensor data;

[0048] Step two, real-time monitoring of dust concentration, triggering compressed air blowing or ultrasonic vibration cleaning when the dust concentration exceeds the set dust concentration threshold; for example: when compressed air blowing cleaning, the compressed air pressure is 0.5-0.8 MPa, delivered to the blowing port through a stainless steel pipe with an inner diameter of 8 mm, the blowing port uses a specially designed fan-shaped nozzle, the blowing angle is 120°, the blowing time is 5-10 seconds, and the blowing frequency is every 2 hours, which can be adjusted according to the actual dust concentration and cleaning effect; before blowing, the measurement circuit of the sensor is closed to prevent electrostatic interference with the measurement data during the blowing process;

[0049] When ultrasonic vibration cleaning, the frequency of the ultrasonic generator is set to 40-60 kHz, the power is 80-120 W, and the vibration time is 3-5 minutes; the ultrasonic transducer is tightly connected to the sensor shell through a specially designed silica gel pad, ensuring efficient transmission of vibration energy to the sensor surface, while avoiding mechanical damage to the sensor.

[0050] After cleaning, perform temperature sensor response test, if the actual time consumption of temperature reaching the set drop difference exceeds the set time consumption, determine that the temperature sensor is abnormal, start the standby temperature sensor and alarm at the same time;

[0051] Step three, fuse the temperature sensor, infrared thermal imager and air conditioner outlet temperature data according to their respective weight coefficients and combine with the humidity sensor data to output the corrected environmental temperature value;

[0052] If the difference between the temperature sensor reading and the corrected ambient temperature value exceeds the preset temperature range interval, the backup temperature sensor is started and an alarm is given.

[0053] In this embodiment, the ambient temperature is monitored by a multi-modal sensor, so that the ambient temperature can be quickly and accurately obtained, reliable temperature data is provided for accurate control of the air conditioner, overcooling of the air conditioner caused by inaccurate temperature measurement is avoided, and energy utilization efficiency is improved; at the same time, through dynamic cleaning, the dust on the temperature sensor is effectively cleaned.

[0054] The step two further comprises:

[0055] A mobile cleaning robot is used to regularly patrol and clean the surface of each temperature sensor along a preset path; the mobile cleaning robot has autonomous navigation function, is equipped with laser radar and visual camera for navigation and environment perception, can patrol and clean the temperature sensor along the preset path, is configured with a brush made of carbon fiber material and a high-suction dust collection device, and can effectively remove dust without scratching the surface of the temperature sensor.

[0056] After cleaning, the change rate of the temperature sensor data before and after cleaning is compared, and if it is lower than the set change rate threshold, it is determined that the cleaning operation is invalid, and manual intervention is notified.

[0057] In the present application, on the basis of basic compressed air blowing and ultrasonic vibration cleaning, a mobile cleaning robot is introduced for regular inspection and cleaning, the robot can reach positions that are not easily reached by manual operation, and the temperature sensor is cleaned comprehensively, further reducing the adhesion of dust on the surface of the temperature sensor, making up for the shortcomings of traditional cleaning methods, and ensuring that the temperature sensor is in good working condition for a long time; at the same time, the cleaning operation effectiveness is judged by comparing the change rate of the temperature sensor data before and after cleaning, and if it is lower than the set threshold, manual intervention is notified, so that incomplete cleaning can be found in time, the cleaning quality is ensured, the temperature sensor continues to report false alarms due to ineffective cleaning is avoided, and the accuracy and reliability of the temperature sensor data are ensured; through the mobile cleaning robot, automatic inspection and cleaning are carried out according to the preset path, the workload of manual regular inspection and cleaning of the temperature sensor is reduced, and the cost of manual maintenance is reduced; at the same time, manual intervention is more targeted, and the maintenance efficiency is improved.

[0058] The method further comprises:

[0059] By using a pre-trained long short-term memory (LSTM) neural network model, the influence trend of dust accumulation on the temperature sensor is predicted, and the air conditioner cooling power is dynamically adjusted.

[0060] In the area where the dust concentration continues to exceed the standard, the weight of the air conditioner outlet temperature is increased, and the weight of the temperature sensor is reduced.

[0061] In this invention, in areas where dust concentrations consistently exceed standards, sensor weights are dynamically adjusted, increasing the weight of air conditioner outlet temperature sensors and decreasing the weight of temperature sensors, which are more susceptible to dust effects. This strategy better adapts to harsh, high-dust environments, accurately reflecting the actual ambient temperature even when sensor data is interfered with. It optimizes air conditioner operation control, avoids energy waste due to sensor errors, and, by predicting dust impacts and optimizing air conditioner operating parameters in advance, reduces the likelihood of frequent erroneous adjustments or operation under harsh conditions, thus lowering equipment wear and tear, extending the lifespan of air conditioning equipment, and reducing maintenance and replacement costs.

[0062] The process of fusing temperature data from temperature sensors, infrared thermal imagers, and air conditioning vents according to their respective weighting coefficients, and combining this data with humidity sensor data, to output the corrected ambient temperature value is as follows:

[0063] After preprocessing the data from the temperature sensor, infrared thermal imager, and air conditioner vent temperature and humidity sensor, dimensionless first temperature values ​​were obtained. Second temperature value Third temperature value Humidity value S;

[0064] Substitute into the formula:

[0065] The corrected ambient temperature value was calculated. ;

[0066] In the formula, ; , They represent the first The maximum and minimum values ​​among the first, second, and third temperature values ​​during the sampling; , , , These are weighting coefficients, determined based on historical data analysis. They are used to assign different weights to different sensor data based on their reliability, stability, and contribution to the final corrected temperature value in the actual environment. For example, in a specific high-dust environment, if the temperature sensor is less affected by dust and its data accuracy is higher, it can be assigned a higher weight. A larger value is chosen to highlight its role in calculating the corrected temperature value; This represents the total number of samples taken over a set time period, counting backwards from the current moment. It is used to synthesize sensor data over a given timeframe. In high-dust environments, sensor data may fluctuate due to factors such as dust adhesion. By sampling multiple times and performing comprehensive calculations, the impact of random errors on the corrected temperature value can be reduced, making the results more stable and reliable. Indicates the first Subsequently, is a judgment function about humidity value, is an adjustment parameter greater than 0, determined based on historical data analysis, which controls the attenuation speed of the exponential function in the denominator position of the exponential term; in a high dust environment, the dispersion degree of sensor data has a greater impact on the accuracy of the corrected temperature value through the range; the greater the R value, the lower the sensitivity of the exponential function to the range; the smaller the R value, the more sensitive the exponential function to the change of the range; adjusting R according to factors such as dust concentration in the actual environment and sensor characteristics can better balance the influence of data dispersion degree on the corrected temperature value. The value range of R is 0.1≤R≤10, and when R is taken in the range, the greater the R value, the lower the sensitivity of the exponential function to the range; for example: when R=0.1, a slight change in the range will cause a large change in the exponential function value; and when R=10, the range needs to have a large change for the exponential function value to have a significant change. the sensitivity of the exponential function to the range is lower; for example: when R=0.1, a slight change in the range will cause a large change in the exponential function value; and when R=10, the range needs to have a large change for the exponential function value to have a significant change.

[0067] In the present application, the data of the temperature sensor, the infrared thermal imager, the air conditioner outlet temperature and the humidity sensor are fused and calculated by the formula:

[0068] The data of the temperature sensor, the infrared thermal imager, the air conditioner outlet temperature and the humidity sensor are fused and calculated by the formula: The data of the temperature sensor, the infrared thermal imager, the air conditioner outlet temperature and the humidity sensor are fused and calculated by the formula: The data of the temperature sensor, the infrared thermal imager, the air conditioner outlet temperature and the humidity sensor are fused and calculated by the formula: The data of the temperature sensor, the infrared thermal imager, the air conditioner outlet temperature and the humidity sensor are fused and calculated by the formula:

[0069] By the technical solution, the multiple sampling data in a period of time are calculated, and the abnormal data are inhibited according to the data dispersion degree, so that the system stability is enhanced, the air conditioner control error caused by abnormal data is reduced, and the stable operation of the air conditioner management system is ensured; the more accurate corrected environment temperature value provides a reliable basis for the operation control of the air conditioner, the air conditioner can more accurately adjust the refrigeration power according to the accurate data, the energy waste or comfort reduction problem caused by temperature misjudgment is avoided, more efficient and intelligent air conditioner control is realized, and the energy saving effect and user comfort of the system are improved.

[0070] The expression of the judgment function about the humidity value is as follows:

[0071]

[0072] In the formula, represents the lower limit of the measured humidity value, represents the pre-set humidity interval limit, represents the proportional coefficient corresponding to different humidity intervals, which is obtained by fitting according to experimental or historical data.

[0073] In the present application, the humidity interval is divided by In the actual high-dust environment, the influence of humidity on temperature measurement is not uniform, different humidity ranges may have different degrees of influence on sensor accuracy, dust adhesion characteristics and human perception of temperature; for example, at a lower humidity, dust is more likely to adhere to the surface of the sensor, which may cause a large temperature measurement deviation; at a higher humidity, the humidity itself may interfere with the heat conduction process of the temperature sensor; by setting these interval limits, different correction strategies can be adopted for different humidity ranges; by using a segmented function, different calculation methods are set for different humidity intervals, which can more accurately consider the influence of humidity on the corrected environment temperature value; compared with using a single function form, this segmented calculation method can better adapt to the complex influence mechanism of humidity on temperature measurement in different ranges, and improve the accuracy of the corrected temperature value; in combination with the calculation formula of the corrected environment temperature value, the data of the temperature sensor, the infrared thermal imager, the air outlet temperature of the air conditioner and the humidity sensor are fused, not only the weights of the sensor data are considered, but also the range information and humidity factors of the data are used to adjust the calculation result, so that the corrected environment temperature value more accurately reflects the actual environment temperature, and the temperature measurement error caused by dust interference of the sensor is effectively reduced. ​​​​​​

[0074] The process of predicting the influence trend of dust accumulation on the temperature sensor by the pre-trained long short-term memory (LSTM) neural network model is as follows:

[0075] The first, second, third temperature values, humidity value, and dust concentration are input into the pre-trained long short-term memory (LSTM) neural network model, and the accumulated dust amount of the temperature sensor per unit time, i.e., the dust accumulation rate G, is output.

[0076] The dust accumulation rate G is compared with the preset dust accumulation threshold G0. If G≥G0, it is predicted that the influence trend of dust accumulation on the temperature sensor is abnormal.

[0077] Otherwise, it is predicted that the influence trend of dust accumulation on the temperature sensor is normal.

[0078] In the present application, the first, second, third temperature values, humidity value, and dust concentration are input into the pre-trained long short-term memory (LSTM) neural network model, and the accumulated dust amount of the temperature sensor per unit time, i.e., the dust accumulation rate G, is output. The dust accumulation rate G is compared with the preset dust accumulation threshold G0. This method can accurately predict the influence trend of dust accumulation on the temperature sensor by using multi-source data and a powerful neural network model. The dust accumulation condition can be known in advance, which provides a basis for timely measures to avoid excessive dust accumulation, which leads to a large measurement error of the sensor and further affects the accurate temperature control of the air conditioner.

[0079] When the influence trend of dust accumulation on the temperature sensor is abnormal, the sensor can be cleaned or maintained in time to prevent the performance of the sensor from declining. Compared with the traditional post-maintenance method, this preventive maintenance can reduce the frequency of sensor failure, reduce the abnormal operation of the air conditioner caused by sensor failure, ensure the stable operation of the system, and prolong the service life of the sensor and the air conditioner.

[0080] The process of dynamically adjusting the refrigeration power of the air conditioner is as follows:

[0081] If G≥G0 and the difference between the temperature sensor reading and the corrected ambient temperature value in a historical time period is in an increasing state, the refrigeration power of the air conditioner is reduced. Otherwise, the current refrigeration power of the air conditioner is maintained.

[0082] In the application, the refrigeration power of the air conditioner is dynamically adjusted according to the dust accumulation rate G and the difference between the temperature sensor reading and the corrected ambient temperature value in the backtracking history time period; when G is greater than or equal to G0 and the difference is in an increasing state, the refrigeration power is reduced, which avoids excessive refrigeration of the air conditioner in the case that the sensor may have errors, reduces energy waste, maintains the indoor temperature in a reasonable range, guarantees the comfort of the user, and realizes the balanced optimization of energy saving and comfort; the way of dynamically adjusting the refrigeration power according to the actual environmental parameter changes enables the air conditioning system to better adapt to the complexity of the high-dust environment, that is, even in the case that the dust concentration changes constantly and the sensor measurement may have errors, the stable operation state can be maintained, and automatic adjustment is made to adapt to the environmental changes.

[0083] The process of judging that the dust concentration is continuously overstandard is as follows:

[0084] The dust concentration average in a history period in the current area is compared with the preset warning value;

[0085] If the dust concentration average is greater than or equal to the preset warning value, and the dust accumulation rate G of at least one temperature sensor in the current area is greater than or equal to G0, it is judged that the current dust concentration is continuously overstandard, and the current area is marked as the area with continuously overstandard dust concentration.

[0086] In the application, the dust concentration average in a history period in the current area is compared with the preset warning value, the dust accumulation rate G of the temperature sensor is combined to judge whether the current dust concentration is continuously overstandard, and the corresponding area is marked; the above method can accurately identify the area with continuously overstandard dust concentration, provides a basis for taking targeted measures, such as strengthening the cleaning frequency of the sensor in the area, adjusting the air conditioner operation mode, and guarantees the normal operation of the air conditioning system in the area; at the same time, the area with continuously overstandard dust concentration is marked and specially processed, which can effectively avoid the air conditioner control failure caused by the dust problem; by adjusting the sensor weight and optimizing the air conditioner operation parameters, it is ensured that the air conditioner in the high-dust area can still refrigerate according to the actual environmental demand, maintains the comfort of the indoor environment, and improves the local operation effect of the air conditioning system in the complex environment.

[0087] It should be noted that the calculation formula and the parameters participating in the operation in the application are all pre-processed by dimensionless processing, and the process of dimensionless processing is known in the industry, which is not described here.

[0088] The above describes one embodiment of the application in detail, but the content described is only the preferred embodiment of the application and cannot be considered as limiting the scope of the implementation of the application. Any equivalent changes and improvements made within the scope of the application should still belong to the patent coverage range of the application.

Claims

1.A method of managing an air conditioner of an Internet of Things, characterized by, The method comprises the following steps: Step one, environment monitoring by multi-modal sensors deployed in high dust area; including temperature sensor, humidity sensor and infrared thermal imager; Step two, real-time monitoring of dust concentration, when the dust concentration exceeds the set dust concentration threshold, triggering compressed air blowing or ultrasonic vibration cleaning; After cleaning, perform response test of temperature sensor, if the actual time consumption of temperature reaching the set drop difference exceeds the set time consumption, determine that the temperature sensor is abnormal, start the backup temperature sensor and alarm at the same time; Step three, fuse temperature sensor, infrared thermal imager and air conditioner outlet temperature data according to their respective weight coefficients, and combine humidity sensor data to output corrected environmental temperature value; If the difference between the temperature sensor reading and the corrected environmental temperature value exceeds the preset temperature range interval, start the backup temperature sensor and alarm; The process of fusing temperature sensor, infrared thermal imager and air conditioner outlet temperature data according to their respective weight coefficients and combining humidity sensor data to output corrected environmental temperature value is: After the data of the temperature sensor, the infrared thermal imager, the air conditioner outlet temperature and the humidity sensor are preprocessed, dimensionless first temperature value , second temperature value , third temperature value and humidity value S are obtained respectively. Substitute the formula: corrected ambient temperature value is calculated ; In the formula, ; , respectively represent the maximum value and the minimum value among the first, second and third temperature values at the i-th sampling time; , , , , is a weight coefficient, which is determined based on historical data analysis; represents the total number of samplings within a set period of time from the current time, represents the i-th, , is a judgment function about the humidity value, is an adjustment parameter greater than 0, which is determined based on historical data analysis; the judging function with respect to the humidity value the expression is: ； In the formula, represents the lower limit of the measured humidity value, , represents the pre-set humidity interval limit, , , represents the proportional coefficient corresponding to different humidity intervals, which is obtained by fitting experimental or historical data. 2.The management method of the air conditioner of the IoT according to claim 1, characterized in that, The step two further comprises the following steps before performing the response test of temperature sensor: Use a mobile cleaning robot to regularly patrol and clean the surface of each temperature sensor along the preset path; After cleaning, compare the change rate of temperature sensor data before and after cleaning, if it is lower than the set change rate threshold, determine that the cleaning operation is invalid, and notify manual intervention. 3.The method of Claim 2, wherein, The method further comprises: Through a pre-trained long short-term memory (LSTM) neural network model, predict the influence trend of dust accumulation on the temperature sensor, and dynamically adjust the refrigeration power of the air conditioner; In the area where the dust concentration continues to exceed the standard, increase the weight of the air conditioner outlet temperature and reduce the weight of the temperature sensor. 4.The method of Claim 1, wherein, The process of predicting the influence trend of dust accumulation on the temperature sensor through the pre-trained long short-term memory (LSTM) neural network model is: Input the first, second, third temperature values, humidity value and dust concentration into the pre-trained long short-term memory (LSTM) neural network model, and output the accumulated dust amount of the temperature sensor per unit time, i.e. the dust accumulation rate G; Compare the dust accumulation rate G with the preset dust accumulation threshold G0, if G≥G0, predict that the influence trend of dust accumulation on the temperature sensor is abnormal; Otherwise, predict that the influence trend of dust accumulation on the temperature sensor is normal. 5.The method of Claim 4, wherein, The process of dynamically adjusting the refrigeration power of the air conditioner is: If G≥G0 and the difference between the temperature sensor reading and the corrected environmental temperature value in a historical time period is increasing, reduce the refrigeration power of the air conditioner; otherwise, maintain the current refrigeration power of the air conditioner. 6.The method of Claim 5, wherein, The process of judging whether the dust concentration continues to exceed the standard is: Compare the average dust concentration in a historical period in the current area with the preset warning value; If the average dust concentration ≥ the preset warning value, and the dust accumulation rate G of at least one temperature sensor in the current area ≥ G0, determine that the current dust concentration continues to exceed the standard, and mark the current area as the area where the dust concentration continues to exceed the standard.

Images