CN-122015413-A - Refrigerator compressor start-stop control method based on data processing

CN122015413ACN 122015413 ACN122015413 ACN 122015413ACN-122015413-A

Abstract

The application relates to the technical field of refrigerator control, in particular to a refrigerator compressor start-stop control method based on data processing, which comprises the steps of collecting multidimensional time sequence data in the refrigerator operation process; the method comprises the steps of responding to a compressor to be switched to a shutdown state, calculating a warming inertia index according to time required by unit temperature rise in a refrigerator, determining attention weight of historical information according to the warming inertia index and a switching frequency, carrying out weighted fusion on the historical information and a short-time state in a long-time memory network based on the attention weight to obtain a fusion feature vector at the current control moment, inputting the fusion feature vector into a prediction regression layer to obtain a prediction temperature, and realizing start-stop control of the compressor according to the prediction temperature. By the technical scheme, the self-adaptability of start-stop control and the stability of the temperature in the box can be improved.

Inventors

Liang Ronghuan
LI JIAJIN
CHEN ZHIFENG
ZHU JIANGHUI

Assignees

佛山市顺德区大磐电器实业有限公司

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (10)

1. The refrigerator compressor start-stop control method based on data processing is characterized by comprising the steps of collecting multidimensional time sequence data in the refrigerator operation process, wherein the multidimensional time sequence data comprise the temperature in the refrigerator, the ambient temperature and the door opening and closing state; Responding to the switching of the compressor to a stop state, and calculating a back-temperature inertia index according to the time required by unit temperature rise in the refrigerator; Counting the switching frequency of a refrigerator door in a preset time window to determine the attention weight of the history information, wherein the attention weight is inversely related to the switching frequency and positively related to the return-temperature thermal inertia index, and the preset time window starts when the compressor is switched to a shutdown state and ends when the compressor is switched to a current control moment; Weighting and fusing the history information of the current control moment and the short-time state in the long-short-time memory network based on the attention weight to obtain a fused feature vector of the current control moment, wherein the history information is an average value of a plurality of short-time states before the current control moment; and inputting the fusion feature vector into a predictive regression layer to obtain the predicted temperature at the future moment, and realizing the start and stop control of the compressor according to the predicted temperature and the temperature in the real-time box.
2. The method for controlling the start and stop of the refrigerator compressor based on data processing according to claim 1 is characterized in that after multi-dimensional time sequence data in the operation process of the refrigerator are collected, the method further comprises the steps of conducting moving average filtering processing on the collected temperature in the refrigerator to remove random noise, responding to the fact that the duration time of the state change of the refrigerator door exceeds a preset threshold value, recording the state change as a door on-off state once, and conducting normalization processing on the processed temperature in the refrigerator, the environment temperature and the door on-off state to obtain input data for inputting a long-short time memory network.
3. The method for controlling start and stop of a refrigerator compressor based on data processing according to claim 1, wherein calculating the back-warming inertia index according to the time required for unit temperature rise in the refrigerator comprises responding to the compressor switching to a stop state, recording the time in the stop state as a starting time and recording the starting time and the temperature in the starting box, continuously monitoring until the current control time is reached, recording the current control time as a cut-off time and recording the cut-off time and the temperature in the cut-off box, calculating the time difference between the cut-off time and the starting time, taking the absolute value of the difference between the cut-off temperature and the temperature in the starting box as an effective temperature rise, and taking the ratio of the time difference and the effective temperature rise as the back-warming inertia index of the current control time.
4. The method for controlling start-stop of a refrigerator compressor based on data processing as claimed in claim 3, wherein the effective temperature rise further comprises calculating a difference between the cut-off temperature and the start temperature as a first temperature difference, calculating an average ambient temperature and an average in-tank temperature between the cut-off time and the start time, taking the difference between the average ambient temperature and the average in-tank temperature as a second temperature difference, subtracting a product of a heat exchange coefficient ratio and the second temperature difference from the first temperature difference, and taking an absolute value of the difference as the effective temperature rise.
5. The method for determining the heat exchange coefficient ratio comprises controlling a refrigerator in an idle state to sequentially operate at least two different preset environmental temperatures, acquiring a natural temperature rise value in a preset calibration time window during the shutdown of the compressor for each preset environmental temperature, calculating a temperature difference average value of an average environmental temperature and an average box temperature in the calibration time window, and calculating a ratio of the natural temperature rise value to the temperature difference average value at each preset environmental temperature, and determining an arithmetic average value of the ratios as the heat exchange coefficient ratio.
6. The method for controlling start and stop of a refrigerator compressor based on data processing according to claim 1, wherein the step of counting the switching frequency of the refrigerator door within the preset time window to determine the attention weight of the history information includes calculating a confidence coefficient which is inversely related to the switching frequency, calculating a difference between the return thermal inertia index and the reference thermal inertia index, dividing the difference by the reference thermal inertia index to obtain a normalized difference, and obtaining the attention weight according to the confidence coefficient and the normalized difference.
7. The method for controlling start and stop of a refrigerator compressor based on data processing as claimed in claim 6, wherein the attention weight is a weight of attention The relation is satisfied: Wherein, the method comprises the steps of, For the current control moment Is used for the temperature return coefficient of inertia (CTI), Is the reference thermal inertia index; is a confidence coefficient.
8. The method for controlling start and stop of a refrigerator compressor based on data processing as set forth in claim 1, wherein the current control time is the same as the current control time Is a fusion feature vector of (1) The relation is satisfied: ; Wherein, the For the current control moment Is used for the concentration weight of the person, For the current control moment Is used to determine the history information of the (c), For the current control moment Is a short time state of (c).
9. The method for controlling start-stop of a refrigerator compressor based on data processing as recited in claim 1, wherein said controlling start-stop of the compressor based on the predicted temperature and the real-time in-tank temperature includes starting the compressor in response to a difference between the predicted temperature and a set temperature being greater than a maximum allowable temperature.
10. The refrigerator compressor start-stop control method based on data processing according to claim 1 is characterized in that the training process of the long-short time memory network comprises the steps of collecting historical operation data of a refrigerator as a training set, taking multi-dimensional time sequence data of historical time as input of the long-short time memory network and actual temperature in the refrigerator at the next future time as a label, calculating mean square error loss between predicted temperature output by the long-short time memory network and the label, and iteratively updating the long-short time memory network by using a back propagation algorithm until the mean square error loss is smaller than a preset convergence threshold or the iteration number is larger than a preset number.

Description

Refrigerator compressor start-stop control method based on data processing Technical Field The application relates to the technical field of refrigerator control, in particular to a refrigerator compressor start-stop control method based on data processing. Background The refrigerator is used as refrigeration equipment for keeping a constant low-temperature environment, and mainly depends on intermittent operation of a compressor to adjust the temperature in the refrigerator, and the control strategy of the refrigerator directly influences the fresh-keeping effect of food materials and the energy consumption level of the equipment. With the development of intelligent technology, intelligent control based on data processing gradually replaces traditional mechanical temperature control. At present, the patent application with the application publication number of CN119665526A discloses a refrigerator and a refrigerator compressor operation control method, wherein the refrigerator comprises a refrigerator body, a compressor, an inter-room temperature sensor, an environment temperature sensor and a controller, wherein a door body is arranged at an opening of the refrigerator body, at least one storage chamber is arranged in the refrigerator body and used for storing objects, the compressor is used for compressing and circulating a refrigerant, the inter-room temperature sensor is used for collecting the inter-room temperature of the storage chamber, and the environment temperature sensor is used for collecting the environment temperature in the room. The method comprises the steps of obtaining the room temperature, the environment temperature and the current set temperature, constructing a feature vector according to the room temperature, the environment temperature and the current set temperature, inputting the feature vector into a pre-trained neural network model for calculation, obtaining working frequency output, and controlling the compressor to run at the working frequency. The method directly inputs a pre-trained neural network model according to the compartment temperature, the environment temperature and the current set temperature acquired in real time, and controls the compressor according to the output working frequency, however, in the operation process of the refrigerator, the change of the internal load state of the refrigerator and the external interference of frequently opening and closing the door can influence the refrigeration effect of the refrigerator, so that the compressor cannot realize self-adaptive control, and the temperature in the refrigerator is unstable. Disclosure of Invention In order to solve the technical problems of low accuracy and low product yield of the start-stop control of the refrigerator compressor based on data processing, the application provides a start-stop control method of the refrigerator compressor based on data processing, which can improve the self-adaptability of the start-stop control of the refrigerator compressor and the stability of the temperature in the refrigerator. The application provides a refrigerator compressor start-stop control method based on data processing, which comprises the steps of collecting multidimensional time sequence data in the running process of a refrigerator, wherein the multidimensional time sequence data comprise temperature in the refrigerator, ambient temperature and a door switch state, responding to the switching of the compressor to a stop state, calculating a return heat inertia index according to time required by unit temperature rise in the refrigerator, counting the switch frequency of the refrigerator door in a preset time window to determine attention weight of historical information, wherein the attention weight is in negative correlation with the switch frequency and in positive correlation with the return heat inertia index, the preset time window starts when the compressor is switched to the stop state and ends when the compressor is in the current control state, weighting and fusing the historical information and the short-time state of the current control moment in a long-short-time memory network based on the attention weight to obtain a fused characteristic vector of the current control moment, the historical information is an average value of a plurality of short-time states before the current control moment, inputting the fused characteristic vector to a prediction layer to obtain a prediction temperature of future moment, and realizing start-stop control of the compressor according to the prediction temperature and the real-time temperature in the refrigerator. The self-adaptive start-stop control of the refrigerator compressor is realized by responding to the condition that the compressor is switched to a stop state to calculate a back-temperature thermal inertia index, determining attention weight according to a gate switching frequency, further carrying out weighted fusion on the history informatio