Methods for regulating air conditioning temperature of subways in smart cities and internet of things systems thereof

1 . A method for regulating air conditioning temperature of a subway in a smart city, comprising:

obtaining, based on at least one collecting device, temperature data and humidity data inside and outside a carriage of a plurality of carriages of the subway, the at least one collecting device including at least one of an infrared temperature sensor, a gas temperature sensor, a radiation temperature sensor, a hair hygrometer, a dry and wet bulb hygrometer, a dew point hygrometer, a coulomb hygrometer, and a temperature and humidity meter; wherein the at least one collecting device includes at least two first collecting devices and at least two second collecting devices, the at least two first collecting devices are installed at a first preset position inside the carriage, and a distance between two adjacent first collecting devices satisfies a first preset condition; and the at least two second collecting devices are installed at a second preset position within a platform, and a distance between two adjacent second collecting devices satisfies a second preset condition;

determining, based on the temperature data and the humidity data, refrigeration power adjustment values and ventilation adjustment values of the plurality of carriages;

generating a control instruction for controlling a temperature regulation device and a ventilation device based on the refrigeration power adjustment values and the ventilation adjustment values of the plurality of carriages, the temperature regulating device including an air conditioner, a refrigerator, and a heater, the ventilation device including a device that directly regulates humidity and a device that indirectly regulates the humidity, the device that directly regulates the humidity including a humidifier and a dehumidifier, the device that indirectly regulates the humidity including an exhaust fan, the control instruction referring to an instruction for controlling an adjustment of a refrigeration power of a subway air conditioner and/or a ventilation of the ventilation device, the control instruction including the refrigeration power adjustment values and the ventilation adjustment values;

controlling the temperature regulation device based on the control instruction to regulate temperature of the plurality of carriages, including:

adjusting current operating power of the temperature regulation device to target operating power according to a corresponding refrigeration power adjustment value of the temperature regulation device; and

controlling the ventilation device based on the control instruction to regulate humidity of the plurality of carriages, including:

adjusting current operating power of the ventilation device to target operating power of the ventilation device according to a corresponding ventilation adjustment value.

2 . The method of claim 1 , wherein the determining, based on the temperature data and the humidity data, refrigeration power adjustment values and ventilation adjustment values of the plurality of carriages includes:

constructing, based on the temperature data and the humidity data, a temperature vector and a humidity vector; and

determining an initial refrigeration power corresponding to the temperature vector and an initial ventilation corresponding to the humidity vector by performing, based on the temperature vector and the humidity vector, a vector matching through a reference database.

3 . The method of claim 2 , wherein the reference database includes a plurality of reference temperature data and a plurality of reference humidity data, and a reference refrigeration power corresponding to each of the plurality of reference temperature data and a reference ventilation corresponding to each of the plurality of reference humidity data.

4 . The method of claim 3 , wherein the reference database further includes reference seasonal data and reference weather data.

5 . The method of claim 2 , further comprising:

predicting, based on the initial refrigeration power and the initial ventilation, temperature change data and humidity change data of the plurality of carriages through a prediction model, wherein the prediction model is a machine learning model; and

determining, based on the temperature change data and the humidity change data, the control instruction, wherein the control instruction includes determining the refrigeration power adjustment values and the ventilation adjustment values.

6 . The method of claim 5 , wherein:

an input of the prediction model includes the temperature data and the humidity data;

the prediction model includes a first extraction layer, a second extraction layer, a first prediction layer, and a second prediction layer;

an input of the first extraction layer includes the temperature data, and an output of the first extraction layer includes a temperature change feature;

an input of the second extraction layer includes the humidity data, and an output of the second extraction layer includes a humidity change feature;

an input of the first prediction layer includes the temperature change feature, the humidity change feature, the initial refrigeration power, and the initial ventilation, and an output of the first prediction layer includes the temperature change data; and

an input of the second prediction layer includes the temperature change feature, the humidity change feature, the initial refrigeration power, and the initial ventilation, and an output of the second prediction layer includes the humidity change data.

7 . The method of claim 5 , wherein the input of the prediction model further includes pedestrian flow data.

8 . The method of claim 7 , wherein the prediction model is obtained through a joint training, and the joint training includes:

taking a plurality of sets of temperature data prior to historical moments, a plurality of sets of humidity data prior to historical moments, historical refrigeration powers, a historical ventilation, and historical pedestrian flow data as training samples;

taking actual temperature change data after a historical moment as a first training label, the first training label being a training label corresponding to the first prediction layer;

taking actual humidity change data after the historical moment as the second training label, the second training label being a training label corresponding to the second prediction layer;

inputting the plurality of sets of temperature data prior to historical moments in the training sample into the first extraction layer to obtain the temperature change feature;

inputting the plurality of sets of humidity data prior to historical moments in the training sample into the second extraction layer to obtain the humidity change feature;

inputting the temperature change feature, the humidity change feature, the historical refrigeration power, the historical ventilation, and the historical pedestrian flow data into the first prediction layer to obtain the output of the first prediction layer;

inputting the temperature change feature, the humidity change feature, the historical refrigeration power, the historical ventilation, and the historical pedestrian flow data into the second prediction layer to obtain the output of the second prediction layer;

inputting the temperature change data output by the first prediction layer and the first training label into a first loss function,

inputting the humidity change data output by the second prediction layer and the second training label into a second loss function;

obtaining a trained prediction model by updating the first prediction layer based on the first loss function and updating the second prediction layer based on the second loss function until a preset condition is satisfied, wherein the preset condition includes that the loss function being less than a threshold or converging, or a training period reaching a threshold.

9 . The method of claim 7 , wherein the pedestrian flow data includes a count of people in the plurality of carriages of the subway, a count of people waiting in front of platforms at a station of the subway, gender and age data of the count of people in the plurality of carriages of the subway or the count of people waiting in front of platforms at the station of the subway, and the pedestrian flow data is used to determine the refrigeration power adjustment values and the ventilation adjustment values of the plurality of carriages and generate the control instruction.

10 . The method of claim 9 , wherein the count of people in the plurality of carriages of the subway and the count of people waiting in front of the platforms at the station of the subway are determined based on a person detection algorithm.

11 . The method of claim 5 , wherein the determining the refrigeration power adjustment values and the ventilation adjustment values includes:

constructing, based on the temperature change data and the humidity change data, a temperature change vector and a humidity change vector; and

determining the refrigeration power adjustment values corresponding to the temperature change vector and the ventilation adjustment values corresponding to the humidity change vector by performing, based on the temperature change vector and the humidity change vector, the vector matching through the reference database.

12 . The method of claim 11 , wherein temperature change vectors refer to vectors corresponding to the temperature change data of a certain subway operation line at a plurality of time points after a collection time point, at least one of the temperature change vectors of the plurality of time points includes the temperature change data of the plurality of carriages of the subway at a corresponding time point; humidity change vectors refer to vectors corresponding to the humidity change data of the certain subway operation line at the plurality of time points after the collection time point, at least one of the humidity change vectors of the plurality of time points includes the humidity change data of the plurality of carriages of the subway at a corresponding time point.

13 . The method of claim 11 , wherein the determining the refrigeration power adjustment values corresponding to the temperature change vector and the ventilation adjustment values corresponding to the humidity change vector by performing, based on the temperature change vector and the humidity change vector, the vector matching through the reference database includes:

determining prediction data based on the temperature change vector and the humidity change vector through the prediction model, wherein the prediction data includes predicted temperature data and predicted humidity data;

determining a target vector based on the prediction data, wherein the target vector refers to a vector corresponding to reference data whose vector distance from the predicted data satisfies vector conditions, the vector conditions include the vector distance from the predicted data is a minimum vector distance or vector distance from the predicted data less than a preset distance threshold, and the reference data is obtained from the reference database;

determining a reference refrigeration power and a reference ventilation corresponding to the target vector as a predicted refrigeration power and a predicted ventilation, wherein the predicted refrigeration power and the predicted ventilation are a refrigeration power of a subway air conditioner corresponding to the predicted temperature data and a ventilation of a ventilation device corresponding to the predicted humidity data;

determining the refrigeration power adjustment values based on a difference between the predicted refrigeration power and the initial refrigeration power; and

determining the ventilation adjustment values based on a difference between the predicted ventilation and the initial ventilation.

14 . The method of claim 2 , wherein the determining the initial refrigeration power corresponding to the temperature vector and the initial ventilation corresponding to the humidity vector by performing, based on the temperature vector and the humidity vector, the vector matching through the reference database includes:

calculating a distance between the temperature vector and a plurality of reference temperature vectors in the reference database, and determining a reference refrigeration power corresponding to a reference temperature vector satisfies a third preset condition as the initial refrigeration power; and

calculating a distance between the humidity vector and a plurality of reference humidity vectors in the reference database, and determining a reference ventilation corresponding to a reference humidity vector satisfies a fourth preset condition as the initial refrigeration power.

15 . The method of claim 1 , wherein

the first preset position includes a position close to an upper side door of the carriage, a position close to a vent, a position away from the vent, and a position above a seat; and

the second preset position includes a position close to a top of a side where the subway stops, a stairway entrance, and an elevator entrance.