Photoinduction electric signal processing method for energy conservation of refrigerator and wine cabinet

Abstract

The invention discloses a light induction electric signal processing method for energy conservation of a refrigerator and a wine cabinet. The method comprises the steps that the opening duration of a cabinet door is marked on a time shaft; obtaining a cabinet door opening angle, internal reference light intensity and external reference light intensity; calculating the internal and external luminescence ratio of the refrigerator wine cabinet; determining a joint correction factor; basic electric energy loss is obtained, and a minimum electric energy loss set value is defined based on the joint correction factor and the cabinet door opening duration; obtaining an electric energy loss prediction value after the refrigerator wine cabinet is opened; comparing the minimum electric energy loss set value with the electric energy loss predicted value to obtain a loss error; if the loss error is greater than a set error threshold value, executing an energy-saving strategy; according to the method, the cabinet door opening angle, duration and nonlinear inflection point angle are considered, so that the electric energy loss of the refrigerator and the wine cabinet can be predicted more accurately, optimization is performed based on the dynamically adjusted minimum electric energy loss set value, and intelligent energy-saving control of the refrigerator and the wine cabinet is realized.

Description

Technical Field

[0001] The invention relates to energy saving of a refrigerator wine cabinet, and in particular to a light sensing electric signal processing method for energy saving of a refrigerator wine cabinet. Background Art

[0002] Wine refrigerators are refrigeration equipment widely used in homes, hotels, and catering industries, mainly for storing and cooling wine. Wine refrigerators usually need to maintain a stable internal environment within a specific temperature range to ensure the quality and taste of the wine. However, in actual use, factors such as frequent opening of the cabinet door or changes in external light often lead to unstable internal temperature of the wine cabinet, which in turn causes increased energy consumption. When the cabinet door is opened, cold air is lost and external heat enters. The compressor and temperature control system of the wine refrigerator must increase the intensity of work to maintain the set temperature, which significantly increases the power loss. At present, the energy-saving control of wine refrigerators mainly relies on traditional temperature control systems and compressor regulation. These systems usually maintain the refrigerated environment inside the wine cabinet by setting a constant temperature, while ignoring the impact of external environmental changes and the door opening state on the internal temperature and energy consumption. Simply relying on temperature control and compressor regulation cannot effectively cope with energy consumption fluctuations in actual applications in some scenarios.

[0003] In addition, most of the energy-saving control methods in the prior art focus on adjusting the temperature or operating frequency, and fail to fully consider the specific impact of the cabinet door opening angle and opening time on the power consumption. In practical applications, the cabinet door opening angle and opening time will directly affect the temperature inside the wine cabinet and the degree of cold air loss. In particular, when the cabinet door opening angle is large, the amount of cold air loss increases, and the speed of external heat entering is accelerated, which will significantly increase the power loss. The current energy-saving control method is usually based on static parameters or empirical data to set the working mode. Although this method is simple and easy to use, it cannot fully adapt to complex and changing environments and usage patterns. For example, the traditional energy-saving solution fails to consider the changes in ambient light when the cabinet door is opened, and the changes in ambient light will affect the stability of the temperature inside the wine cabinet, thereby indirectly increasing energy consumption. Therefore, the existing energy-saving technology has major defects, especially when dealing with factors such as frequent cabinet door opening, external light changes, and differences in usage patterns, the effect is poor. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a light-sensing electrical signal processing method for energy saving in a refrigerator wine cabinet. By real-time monitoring of the cabinet door opening status and ambient light changes, the minimum power loss setting value is dynamically adjusted to achieve more accurate and flexible energy-saving control.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0006] A method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet, comprising:

[0007] Marking the door opening time of the refrigerator wine cabinet on the time axis; wherein the door opening time is generated by the changing light sensing electrical signal in the refrigerator wine cabinet;

[0008] Obtaining a door opening angle, an internal reference light intensity, and an external reference light intensity of the refrigerator wine cabinet within the door opening time;

[0009] Calculating the internal and external luminous ratio of the refrigerator wine cabinet according to the cabinet door opening angle, the internal reference light intensity and the external reference light intensity;

[0010] Determine a joint correction factor according to the internal and external luminous ratio and the door opening angle;

[0011] Obtain the basic power consumption of the refrigerator wine cabinet when the cabinet door is closed, and define the minimum power consumption setting value of the refrigerator wine cabinet during the cabinet door opening time based on the joint correction factor and the cabinet door opening time

[0012] The basic power loss, internal and external luminous ratio, cabinet door opening angle and cabinet door opening time are input into the pre-trained loss prediction model after being standardized, and the power loss prediction value after the refrigerator wine cabinet is opened is output;

[0013] Comparing the minimum power loss setting value with the power loss prediction value to obtain a loss error;

[0014] If the loss error is greater than the set error threshold, the energy-saving strategy is executed;

[0015] The energy-saving strategy reduces the basic power consumption of the refrigerator wine cabinet by adjusting several electrical parameters of the refrigerator wine cabinet to maintain the minimum power consumption setting value.

[0016] In some embodiments, marking the door opening time of the refrigerator wine cabinet on the time axis includes:

[0017] S1-1, preset a dynamic time window with a fixed length on the time axis and sliding with the current time point; wherein the end time point of the dynamic time window is always aligned with the current time point;

[0018] S1-2, collecting the internal light intensity of the refrigerator wine cabinet at consecutive time points within the dynamic time window;

[0019] S1-3, calculating the internal illumination change value at adjacent time points according to the internal illumination intensity at the consecutive time points;

[0020] S1-4, if the internal illumination change value at the adjacent time points changes from negative to positive, then mark the current time point as the initial light change time point;

[0021] S1-5, after the initial light change time point, continuously calculating the internal illumination change value at adjacent time points until the internal illumination change value changes from positive to negative, and marking the current time point as a secondary light change time point;

[0022] S1-6. Define the time difference between the initial light change time point and the secondary light change time point as the door opening duration.

[0023] In some embodiments, the internal and external luminous intensity ratio of the refrigerator wine cabinet is calculated according to the cabinet door opening angle, the internal reference light intensity and the external reference light intensity, including:

[0024] S3-1, predefined angle correction constant and door shielding coefficient;

[0025] S3-2, determining a first angle correction factor according to the angle correction constant and the sine value of the door opening angle;

[0026] S3-3, determining an occlusion correction factor according to the cabinet door occlusion coefficient and the sine value of the cabinet door opening angle;

[0027] S3-4, determining the external corrected light intensity according to the first angle correction factor and the shading correction factor;

[0028] S3-5. Define the ratio of the internal reference light intensity to the corrected light intensity as the internal-external luminescence ratio.

[0029] The definition expression of the internal and external luminous ratio is:

[0030]

[0031] Among them, R in / out I is the internal and external luminous intensity ratio, which represents the ratio of the internal reference light intensity to the external corrected light intensity; ri I is the internal reference light intensity, which represents the reference luminous intensity of the internal light source; so is the external reference light intensity, which characterizes the reference luminous intensity of the external light source; sin(θ)+c1 is the first angle correction factor, θ represents the door opening angle, c1 represents the predefined angle correction constant, which avoids sin(θ)+c1 tending to 0 at small angles; 1-k·sin(θ) is the shading correction factor, k represents the door shading coefficient, which characterizes the proportion of light shading by the door frame.

[0032] In some embodiments, determining a joint correction factor according to the internal-external luminous ratio and the door opening angle includes:

[0033] S4-1. Predefine the inflection angle of the door opening angle and the nonlinear growth rate of the inflection angle

[0034] S4-2, extracting a first angle correction factor, and determining a second angle correction factor based on the inflection point angle and its nonlinear growth rate;

[0035] The expression of the second angle correction factor is:

[0036]

[0037] Among them, θ0 represents the inflection point angle of the cabinet door opening angle, β represents the nonlinear growth rate after the cabinet door opening angle exceeds the inflection point angle, Φ θ The second angle correction factor representing the comprehensive second angle correction factor, the inflection point angle and the nonlinear growth rate is used for nonlinear correction of the influence of the cabinet door opening angle exceeding the inflection point angle on the light entry and heat loss;

[0038] S4-3, performing a weighted summation of the second angle correction factor and the internal and external luminous ratio, and then performing an exponential operation with a natural base to generate the joint correction factor;

[0039] The expression of the joint correction factor is:

[0040] Φ j =1+exp(w1·R in / out +w2·Φ θ );

[0041] Among them, Φ j represents the joint correction factor, w1 represents the weight of the internal and external luminous ratio, and w2 represents the weight of the second angle correction factor.

[0042] In some embodiments, the definition expression of the minimum power loss setting value is:

[0043] E o =E b ·Φ j ·T=E b ·(1+exp(w1·R in / out +w2·Φ θ ))·T;

[0044] Among them, E o Indicates the defined minimum power loss setting value, E b It is the basic power consumption of the refrigerator wine cabinet when the cabinet door is closed.

[0045] T represents the door opening time during which the door is opened to maintain the door opening angle.

[0046] In some of the embodiments, the pre-training step of the loss prediction model includes:

[0047] S6-1. Obtain a loss sample training set; wherein the input feature of the loss sample is defined as a simulated state vector constructed by the cabinet door opening angle, the cabinet door opening time, the internal reference light intensity, and the external reference light intensity, and the target label is defined as the actual power loss corresponding to the simulated state vector;

[0048] S6-2. Input the loss sample training set into a general neural network model, and generate the loss prediction model after supervised learning iterative training.

[0049] In some of the embodiments, obtaining a loss sample training set includes:

[0050] S6-1-1, preset the cabinet door opening angle D1, cabinet door opening time T1;

[0051] S6-1-2. Under the simulation state of maintaining the external reference light intensity unchanged, open the cabinet door so that the cabinet door opening angle is D1, the cabinet door opening time is T1, and measure the actual power loss SE1 of the simulation state;

[0052] S6-1-3, construct the external reference light intensity, the cabinet door opening angle D1, the cabinet door opening time T1 and the rated internal reference light intensity as the simulation state vector, and define the actual power loss SE1 as the target label of the simulation state vector;

[0053] S6-1-4, gradually increase the cabinet door opening angle and cabinet door opening time in an arithmetic progression, and repeat S6-1-2 to generate K different simulation state vectors and corresponding target labels;

[0054] S6-1-5, combining the generated simulation state vector and the target label to obtain the loss sample;

[0055] S6-1-6. Arrange the simulation state vectors and the corresponding target labels in a gradually increasing order to construct the loss sample training set.

[0056] In some embodiments, the loss sample training set is input into a general neural network model, and after supervised learning iterative training, the loss prediction model is generated, including:

[0057] S6-2-1. Extract the first batch of loss samples and use them as input to the general neural network model;

[0058] S6-2-2, according to the loss samples of the first batch, perform forward propagation on them, and output the predicted value of the power loss of the first batch in the output layer;

[0059] S6-3-3, taking the mean square error between the predicted value of the first batch of power loss and the actual power loss of the first batch as the target loss of the first batch;

[0060] S6-4-4. Calculate the gradient of the target loss of the first batch relative to the current model parameters;

[0061] S6-4-5, updating the model parameters according to the calculated gradient using the gradient descent method to generate a real-time parameter model;

[0062] S6-4-6, extract the next batch of loss samples, use them as the model input of the real-time parameter model, and repeat S6-2-2 to S6-2-5 until the output target loss is minimized;

[0063] S6-2-7. When the target loss is minimized, the real-time parameter model is defined as the loss prediction model.

[0064] The present invention introduces the cabinet door opening state (including the cabinet door opening angle and opening time) as a key factor affecting power consumption, and further adjusts the nonlinear angle inflection point based on this. Based on the minimum power loss setting value that changes dynamically with different cabinet door opening states, more accurate energy-saving control can be achieved. The specific beneficial effects are as follows:

[0065] The present invention combines the cabinet door opening angle and opening duration to dynamically calculate the minimum power loss setting value based on different opening states. In particular, by introducing a nonlinear inflection point angle, when the cabinet door opening angle reaches a certain critical point, the power loss will increase sharply in a nonlinear manner. Therefore, the power loss target can be flexibly adjusted, and the energy efficiency can be automatically optimized when the door opening state changes, which can better adapt to the changing usage environment.

[0066] Furthermore, the present invention generates a large number of training samples under "different opening states" by simulating state vectors. It does not need to rely on actual historical data, but generates a large number of samples by simulating various factors such as ambient lighting, cabinet door opening angle and duration. This enables the model to be trained based on different opening states, avoiding the data shortage problem caused by the lack of diversified historical data in traditional methods. Through this simulation method, the system can quickly obtain diversified training data and improve the generalization ability and prediction accuracy of the model.

[0067] In summary, the present invention solves the problem of insufficient response to the door opening state and ambient light changes in the prior art. By introducing the door opening angle, duration, and nonlinear inflection point angle, the power loss of the refrigerator wine cabinet can be predicted more accurately, and optimized based on the dynamically adjusted minimum power loss setting value, significantly improving the energy saving effect. At the same time, by generating rich training samples through simulation data, the defects of the traditional method of lack of diversified data and poor real-time adaptability are further solved. Finally, this solution based on real-time dynamic adjustment and simulation data training realizes intelligent energy-saving control of refrigerator wine cabinets. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a schematic flow chart of a method for processing light-sensing electrical signals for energy saving in a refrigerator and wine cabinet according to the present invention;

[0069] Figure 2 This is a schematic diagram of the process of obtaining the cabinet door opening time according to the present invention;

[0070] Figure 3 The figure is a schematic diagram of the process of constructing the loss sample training set according to the present invention. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] See also Figures 1 to 3 The present invention provides a method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet, the processing method comprising:

[0073] Step S1, marking the door opening time of the refrigerator wine cabinet on the time axis; wherein the door opening time is generated by the changing light sensing electrical signal in the refrigerator wine cabinet;

[0074] The step S1 specifically includes:

[0075] S1-1, preset a dynamic time window with a fixed length on the time axis and sliding with the current time point; wherein the end time point of the dynamic time window is always aligned with the current time point;

[0076] S1-2, collecting the internal light intensity of the refrigerator wine cabinet at consecutive time points within the dynamic time window;

[0077] S1-3, calculating the internal illumination change value at adjacent time points according to the internal illumination intensity at the consecutive time points;

[0078] S1-4, if the internal illumination change value at the adjacent time points changes from negative to positive, then mark the current time point as the initial light change time point;

[0079] S1-5, after the initial light change time point, continuously calculating the internal illumination change value at adjacent time points until the internal illumination change value changes from positive to negative, and marking the current time point as a secondary light change time point;

[0080] S1-6. Define the time difference between the initial light change time point and the secondary light change time point as the door opening duration.

[0081] Among them, the internal light intensity and the internal light change value are both light intensity electrical signals acquired in real time by the photoelectric sensing components in the refrigerator wine cabinet, representing the impact of light source changes on the electrical signals.

[0082] This embodiment collects the internal light intensity at each time point in the wine cabinet in real time by presetting a dynamic time window with a fixed length on the time axis that slides with the current time point. By calculating the light changes at adjacent time points, the changes in light intensity are dynamically captured. When the light change changes from negative to positive, it means that the cabinet door starts to be opened, and the system marks it as the initial light change time point. Subsequently, the system continues to monitor the light change until the light change changes from positive to negative, which is marked as the secondary light change time point, and the cabinet door opening duration is the time difference between the two time points. By dynamically tracking the changes in light intensity, the calculation of the cabinet door opening duration is realized, thereby combining the light intensity change and the time dimension, providing a new duration calculation method that unifies the parameters of the two dimensions of light and time into the same framework.

[0083] The processing method also includes:

[0084] Step S2, obtaining the door opening angle, internal reference light intensity and external reference light intensity of the refrigerator wine cabinet within the door opening time;

[0085] The internal reference light intensity refers to the rated light intensity of the light source inside the wine cabinet when the cabinet door is open. It represents the preset intensity of the light inside the wine cabinet under normal working conditions; the light source inside the wine cabinet may be affected by environmental factors (such as the type of light source, the reflective surface inside the wine cabinet, etc.), and the internal reference light intensity is used as a constant to measure the standard intensity of the internal light source. For example, the light intensity of a certain wine cabinet is 1000 lumens, which is its "internal reference light intensity".

[0086] External reference light intensity refers to the light intensity of the external environment of the wine cabinet, which is used to standardize the light impact of the external environment. It is usually set as a fixed reference value based on the environment of the refrigerator wine cabinet, representing the standard of external light intensity in a specific scene.

[0087] The processing method also includes:

[0088] Step S3, calculating the internal and external luminous ratio of the refrigerator wine cabinet according to the cabinet door opening angle, the internal reference light intensity and the external reference light intensity;

[0089] The step S3 specifically includes:

[0090] S3-1, predefined angle correction constant and door shielding coefficient;

[0091] S3-2, determining a first angle correction factor according to the angle correction constant and the sine value of the door opening angle;

[0092] S3-3, determining an occlusion correction factor according to the cabinet door occlusion coefficient and the sine value of the cabinet door opening angle;

[0093] S3-4, determining the external corrected light intensity according to the first angle correction factor and the shading correction factor;

[0094] S3-5. Define the ratio of the internal reference light intensity to the corrected light intensity as the internal-external luminescence ratio.

[0095] The definition expression of the internal and external luminous ratio is:

[0096]

[0097] Among them, R in / out I is the internal and external luminous intensity ratio, which represents the ratio of the internal reference light intensity to the external corrected light intensity; ri I is the internal reference light intensity, which represents the reference luminous intensity of the internal light source; so is the external reference light intensity, which characterizes the reference luminous intensity of the external light source; sin(θ)+c1 is the first angle correction factor, θ represents the door opening angle, c1 represents the predefined angle correction constant, which avoids sin(θ)+c1 tending to 0 at small angles; 1-k·sin(θ) is the shading correction factor, k represents the door shading coefficient, which characterizes the proportion of light shading by the door frame.

[0098] This embodiment introduces a predefined angle correction constant and a cabinet door shielding coefficient, which are used to correct the lighting difference caused by the change of the cabinet door opening angle and consider the shielding effect of the cabinet door frame on the light. Through the above-mentioned first angle correction factor and shielding correction factor, an adjusted external corrected light intensity is determined, that is, the actual external light intensity after considering the cabinet door opening angle and the shielding effect. Finally, the internal reference light intensity is compared with the external corrected light intensity to obtain the internal and external luminous ratio. The dynamically calculated internal and external luminous ratio can accurately reflect the changes in the internal and external lighting environment when the cabinet door is open. As an important indicator for measuring lighting changes, it helps to predict power loss more accurately.

[0099] The processing method also includes:

[0100] Step S4, determining a joint correction factor according to the internal and external luminous ratio and the cabinet door opening angle;

[0101] The step S4 specifically includes:

[0102] S4-1. Predefine the inflection angle of the door opening angle and the nonlinear growth rate of the inflection angle

[0103] S4-2, extracting a first angle correction factor, and determining a second angle correction factor based on the inflection point angle and its nonlinear growth rate;

[0104] The expression of the second angle correction factor is:

[0105]

[0106] Among them, θ0 represents the inflection point angle of the cabinet door opening angle, β represents the nonlinear growth rate after the cabinet door opening angle exceeds the inflection point angle, Φ θ The second angle correction factor, which represents the comprehensive second angle correction factor, inflection point angle and nonlinear growth rate, is used for nonlinear correction of the influence of the cabinet door opening angle on light entry and heat loss after exceeding the inflection point angle; specifically, the nonlinear growth rate describes the nonlinear image of the cabinet door opening angle on power loss after exceeding the inflection point angle. As the opening angle increases, external heat inflow and cold air loss intensify, resulting in an exponentially rapid increase in power loss; the larger the value, the more drastic the increase in power loss caused by the angle change; the nonlinear growth rate β simulates the influence of the opening angle on the external heat inflow and internal cold air loss; when the door is opened at a larger angle, the nonlinear growth rate β nonlinearly adjusts the rapid increase in loss through the exponential term of the natural base e; and the inflection point angle determines the critical point of the opening angle on the increase in power loss; when the cabinet door opening angle θ is close to the inflection point angle θ0, the growth of power loss changes from slow to fast. is accelerated; it reflects the critical state of the transition of the cabinet door opening angle from "low impact" to "high impact"; for example, when θ is less than θ0, the power loss increases more slowly, and when θ is greater than θ0, the power loss increases rapidly; for the nonlinear growth rate, the power loss corresponding to different opening angles (such as 0°, 15°, 30°, etc.) can be measured, and the formula can be used to fit the best nonlinear growth rate; for the inflection point angle, different opening angles are set, the growth trend of the power loss is recorded, the position where the power loss growth rate changes significantly is determined, and it is used as the inflection point angle.

[0107] S4-3, performing a weighted summation of the second angle correction factor and the internal and external luminous ratio, and then performing an exponential operation with a natural base to generate the joint correction factor;

[0108] The expression of the joint correction factor is:

[0109] Φ j =1+exp(w1·R in / out +w2·Φ θ );

[0110] Among them, Φ j represents the joint correction factor, w1 represents the weight of the internal and external luminous ratio, and w2 represents the weight of the second angle correction factor.

[0111] This embodiment defines the second angle correction factor by the door opening angle, inflection point angle and nonlinear growth rate, and uses the second angle correction factor and the acquisition of the internal and external luminous ratio to combine the correction factor, thereby comprehensively considering the nonlinear influence and multi-dimensional joint influence of the door opening angle and internal and external lighting conditions on the power loss.

[0112] The processing method also includes:

[0113] Step S5, obtaining the basic power consumption of the refrigerator wine cabinet when the cabinet door is closed, and defining the minimum power consumption setting value of the refrigerator wine cabinet during the cabinet door opening time based on the joint correction factor and the cabinet door opening time;

[0114] The definition expression of the minimum power loss setting value is:

[0115] E o =E b ·Φ j ·T=E b ·(1+exp(w1·R in / out +w2·Φ θ ))·T;

[0116] Among them, E o Indicates the defined minimum power loss setting value, E b It is the basic power consumption of the refrigerator wine cabinet when the cabinet door is closed.

[0117] T represents the door opening time during which the door is opened to maintain the door opening angle.

[0118] This embodiment comprehensively considers factors such as the cabinet door opening angle, the internal and external light ratio, and the cabinet door opening time, and finally defines a minimum power loss setting value for the refrigerator wine cabinet for the above state (opening angle, opening time, internal and external light conditions). Among them, the minimum power loss setting value is a dynamically adjusted parameter, which aims to predict and optimize the energy consumption of the refrigerator wine cabinet based on multiple influencing factors (such as cabinet door opening angle, internal and external light ratio, cabinet door opening time, etc.).

[0119] Specifically, the basic power loss belongs to the basic energy consumption of the refrigerator wine cabinet when the cabinet door is closed. It represents the minimum energy consumption level of the wine cabinet when there is no external interference and provides an initial energy consumption benchmark. The internal and external luminous ratio reflects the relative relationship between the light intensity inside and outside the wine cabinet. The internal and external luminous ratio takes into account the cabinet door opening angle and the change of the light source (including the light intensity of the external environment and the intensity of the internal light source). When the cabinet door is open, external light may enter the wine cabinet and change the internal lighting conditions. The higher this ratio is, the greater the impact of the external light source on the internal light intensity, and the temperature and humidity in the wine cabinet may also be affected by more external environmental factors, resulting in increased power loss. The cabinet door opening angle directly affects the heat loss and cold air leakage inside the wine cabinet. A larger opening angle means that more heat may flow in from the external environment, and the cold air will be lost faster, resulting in increased power loss. The door opening time is another factor that affects energy consumption. The longer the cabinet door is open, the greater the temperature change in the wine cabinet, causing the system to require more energy to return to the set temperature. Therefore, the longer the opening time, the greater the power loss.

[0120] In summary, through these parameters, the minimum power consumption setting value is a dynamic target. By setting the sensitivity of the internal and external light ratio and the door opening angle to energy consumption through various weights, and based on the above factors, the power consumption is flexibly adjusted to ensure that the power loss is always kept near the minimum value under different operating conditions, thereby optimizing the energy use of the refrigerator wine cabinet.

[0121] The processing method also includes:

[0122] Step S6, standardize the basic power loss, internal and external luminous ratio, cabinet door opening angle and cabinet door opening time, and input them into the pre-trained loss prediction model, and output the power loss prediction value after the refrigerator wine cabinet is opened;

[0123] The step S6 specifically includes:

[0124] S6-1. Obtain a loss sample training set; wherein the input feature of the loss sample is defined as a simulated state vector constructed by the cabinet door opening angle, the cabinet door opening time, the internal reference light intensity, and the external reference light intensity, and the target label is defined as the actual power loss corresponding to the simulated state vector;

[0125] By collecting and combining the above input features, the power consumption of the refrigerator wine cabinet under different usage conditions can be simulated, and the corresponding actual power consumption can be generated as the target label.

[0126] S6-2. Input the loss sample training set into a general neural network model, and generate the loss prediction model after supervised learning iterative training.

[0127] Furthermore, the step S6-1 specifically includes:

[0128] S6-1-1, preset the cabinet door opening angle D1, cabinet door opening time T1;

[0129] S6-1-2. Under the simulation state of maintaining the external reference light intensity unchanged, open the cabinet door so that the cabinet door opening angle is D1, the cabinet door opening time is T1, and measure the actual power loss SE1 of the simulation state;

[0130] S6-1-3, construct the external reference light intensity, the cabinet door opening angle D1, the cabinet door opening time T1 and the rated internal reference light intensity as the simulation state vector, and define the actual power loss SE1 as the target label of the simulation state vector;

[0131] S6-1-4, gradually increase the cabinet door opening angle and cabinet door opening time in an arithmetic progression, and repeat S6-1-2 to generate K different simulation state vectors and corresponding target labels;

[0132] S6-1-5, combining the generated simulation state vector and the target label to obtain the loss sample;

[0133] S6-1-6. Arrange the simulation state vectors and the corresponding target labels in a gradually increasing order to construct the loss sample training set.

[0134] This embodiment can generate a variety of different usage scenario data online by simulating different usage conditions and environmental conditions (such as cabinet door opening angle, opening time, internal and external light intensity, etc.) without waiting for historical data in actual use. This provides rich scenario data for model training, making the loss prediction model more robust.

[0135] Specifically, the generation of simulation state vectors does not rely on historical data of actual equipment operation, but rather on accurately simulating the working state of the environment and equipment to cover various possible usage conditions, including different lighting, temperature, door opening status, etc., thereby improving the diversity and representativeness of the training set.

[0136] In practical applications, historical data may not be sufficient or representative, especially in new devices or new environments. This problem can be avoided by simulating state vectors. Even in the absence of real data, the system can still generate the required training samples through simulation, thereby maintaining the continuity and effectiveness of model training.

[0137] Therefore, this method of generating training sets by simulating state vectors not only improves the efficiency of data collection, but also enables the model to quickly adapt to different usage environments and conditions, effectively solving the training bottleneck caused by insufficient historical data in traditional methods.

[0138] Furthermore, the step S6-2 specifically includes:

[0139] S6-2-1. Extract the first batch of loss samples and use them as input to the general neural network model;

[0140] S6-2-2, according to the loss samples of the first batch, perform forward propagation on them, and output the predicted value of the power loss of the first batch in the output layer;

[0141] S6-2-3, taking the mean square error between the predicted value of the first batch of power loss and the actual power loss of the first batch as the target loss of the first batch;

[0142] S6-2-4. Calculate the gradient of the target loss of the first batch relative to the current model parameters;

[0143] S6-2-5, updating the model parameters according to the calculated gradient using the gradient descent method to generate a real-time parameter model;

[0144] In this embodiment, the Adam optimization algorithm can be used for parameter update in gradient descent. The Adam algorithm combines momentum and adaptive learning rate to effectively handle the instability in the gradient descent process and ensure that the model parameters converge to the optimal value in iteration. The gradient is calculated through each batch of loss samples, the weights and biases in the neural network are updated, and the target loss is gradually minimized.

[0145] S6-2-6, extract the next batch of loss samples, use them as the model input of the real-time parameter model, and repeat S6-2-2 to S6-2-5 until the output target loss is minimized;

[0146] S6-2-7. When the target loss is minimized, the real-time parameter model is defined as the loss prediction model.

[0147] This embodiment optimizes the prediction accuracy of power loss through supervised learning and iterative training. First, data is extracted from the loss sample training set, and forward propagation is performed to calculate the loss prediction value. The model learns the complex relationship between input and output. By calculating the mean square error between the predicted value and the actual power loss, the model parameters are updated using the gradient descent method to ensure that the model is more accurate after each iteration and eventually converges to the optimal state. This method gradually improves the prediction accuracy based on multi-batch training, reduces the dependence on historical data, and ultimately provides accurate power loss prediction for refrigerator wine cabinet energy-saving control.

[0148] The processing method also includes:

[0149] Step S7, comparing the minimum power loss setting value with the power loss prediction value to obtain a loss error;

[0150] The loss error means the difference between the minimum power loss setting value and the actual predicted power loss. Specifically, the loss error reflects the deviation between the power loss predicted by the current model and the ideal minimum power loss value. A larger loss error means that the electrical parameters of the power loss are not adapted, and the electrical parameters may need to be adjusted to further optimize energy efficiency. A smaller loss error means that the power consumption of the refrigerator is close to the minimum setting value, and the system is already energy-efficient during operation.

[0151] The processing method also includes:

[0152] Step S8: If the loss error is greater than the set error threshold, an energy-saving strategy is executed; wherein the energy-saving strategy reduces the basic power loss of the refrigerator wine cabinet by adjusting several electrical parameters of the refrigerator wine cabinet to maintain the minimum power loss setting value.

[0153] The electrical parameters can be adjusted by:

[0154] Adjust the operating frequency of the compressor to adapt to the current power consumption level;

[0155] Adjust the working mode of the temperature control system to reduce unnecessary energy consumption;

[0156] Optimize the use of other electrical loads, such as interior lighting and fans, and adjust load operation according to real-time demand.

[0157] This embodiment combines photoelectric sensing with an energy loss prediction model to achieve intelligent energy-saving control of a refrigerator wine cabinet. It accurately predicts energy loss by real-time monitoring and analysis of the door opening time, door opening angle, and ambient light changes (the ratio of internal light intensity to external light intensity, i.e., internal and external light intensity ratio), and then optimizes and adjusts the working state of the refrigerator wine cabinet based on the predicted value.

[0158] Specifically, this embodiment defines the minimum power loss setting value by marking the door opening time on the time axis, collecting light intensity through the photoelectric sensing component, and combining the door opening angle. By comparing it with the pre-trained loss prediction model, it can monitor the power loss of the refrigerator wine cabinet in real time, and ensure that the power loss is always close to the optimal setting value by adjusting the operating frequency of the compressor, the temperature control system and other electrical loads, thereby effectively achieving the energy saving goal.

[0159] The above embodiments can be implemented in whole or in part by software, hardware, firmware or other arbitrary combinations. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center.

[0160] The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0161] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division of a waterway underwater terrain change analysis system and method. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0162] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for processing light-sensing electrical signals for energy saving in refrigerators and wine cabinets, characterized in that: include: Marking the door opening time of the refrigerator wine cabinet on the time axis; wherein the door opening time is generated by the changing light sensing electrical signal in the refrigerator wine cabinet; Obtaining the door opening angle, internal reference light intensity, and external reference light intensity of the refrigerator wine cabinet within the door opening time; Calculating the internal and external luminous ratio of the refrigerator wine cabinet according to the cabinet door opening angle, the internal reference light intensity and the external reference light intensity; Determine a joint correction factor according to the internal and external luminous ratio and the door opening angle; Obtaining the basic power consumption of the refrigerator wine cabinet when the cabinet door is closed, and defining the minimum power consumption setting value of the refrigerator wine cabinet within the cabinet door opening time based on the joint correction factor and the cabinet door opening time; The basic power loss, internal and external luminous ratio, cabinet door opening angle and cabinet door opening time are input into the pre-trained loss prediction model after being standardized, and the power loss prediction value after the refrigerator wine cabinet is opened is output; Comparing the minimum power loss setting value with the power loss prediction value to obtain a loss error; If the loss error is greater than the set error threshold, the energy-saving strategy is executed; The energy-saving strategy reduces the basic power consumption of the refrigerator wine cabinet by adjusting several electrical parameters of the refrigerator wine cabinet to maintain the minimum power consumption setting value.

2. The method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet according to claim 1, characterized in that: The door opening time of the refrigerator wine cabinet is marked on the time axis, including: S1-1, preset a dynamic time window with a fixed length on the time axis and sliding with the current time point; wherein the end time point of the dynamic time window is always aligned with the current time point; S1-2, collecting the internal light intensity of the refrigerator wine cabinet at consecutive time points within the dynamic time window; S1-3, calculating the internal illumination change value at adjacent time points according to the internal illumination intensity at the consecutive time points; S1-4, if the internal illumination change value at the adjacent time points changes from negative to positive, then mark the current time point as the initial light change time point; S1-5, after the initial light change time point, continuously calculating the internal illumination change value at adjacent time points until the internal illumination change value changes from positive to negative, and marking the current time point as a secondary light change time point; S1-6. Define the time difference between the initial light change time point and the secondary light change time point as the door opening duration.

3. The method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet according to claim 1, characterized in that: Calculating the internal and external luminous intensity ratio of the refrigerator wine cabinet according to the cabinet door opening angle, the internal reference light intensity and the external reference light intensity, including: S3-1, predefined angle correction constant and door shielding coefficient; S3-2, determining a first angle correction factor according to the angle correction constant and the sine value of the door opening angle; S3-3, determining an occlusion correction factor according to the cabinet door occlusion coefficient and the sine value of the cabinet door opening angle; S3-4, determining the external corrected light intensity according to the first angle correction factor and the shading correction factor; S3-5, defining the ratio of the internal reference light intensity to the corrected light intensity as the internal-external luminescence ratio; The definition expression of the internal and external luminous ratio is: Among them, R in / out I is the internal and external luminous intensity ratio, which represents the ratio of the internal reference light intensity to the external corrected light intensity; ri I is the internal reference light intensity, which represents the reference luminous intensity of the internal light source; so is the external reference light intensity, which characterizes the reference luminous intensity of the external light source; sin(θ)+c1 is the first angle correction factor, θ represents the door opening angle, c1 represents the predefined angle correction constant, which avoids sin(θ)+c1 tending to 0 at small angles; 1-k·sin(θ) is the shading correction factor, k represents the door shading coefficient, which characterizes the proportion of light shading by the door frame.

4. The method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet according to claim 1, characterized in that: According to the internal and external luminous ratio and the cabinet door opening angle, a joint correction factor is determined, including: S4-1. Predefine the inflection angle of the door opening angle and the nonlinear growth rate of the inflection angle S4-2, extracting a first angle correction factor, and determining a second angle correction factor based on the inflection point angle and its nonlinear growth rate; The expression of the second angle correction factor is: Among them, θ0 represents the inflection point angle of the cabinet door opening angle, β represents the nonlinear growth rate after the cabinet door opening angle exceeds the inflection point angle, Φ θ The second angle correction factor representing the comprehensive second angle correction factor, the inflection point angle and the nonlinear growth rate is used for nonlinear correction of the influence of the cabinet door opening angle exceeding the inflection point angle on the light entry and heat loss; S4-3, performing a weighted summation of the second angle correction factor and the internal and external luminous ratio, and then performing an exponential operation with a natural base to generate the joint correction factor; The expression of the joint correction factor is: Φ j =1+exp(w1·R in / out +w2·Φ θ ); Among them, Φ j represents the joint correction factor, w1 represents the weight of the internal and external luminous ratio, and w2 represents the weight of the second angle correction factor.

5. The method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet according to claim 1, characterized in that: The definition expression of the minimum power loss setting value is: From o =E b ·Φ j ·T=E b ·(1+exp(w1·R in / out +w2·Φ θ ))·T; Among them, E o Indicates the defined minimum power loss setting value, E b It is the basic power consumption of the refrigerator wine cabinet when the cabinet door is closed; T represents the door opening time during which the door is opened to maintain the door opening angle.

6. The method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet according to claim 1, characterized in that: The pre-training step of the loss prediction model includes: S6-1. Obtain a loss sample training set; wherein the input feature of the loss sample is defined as a simulated state vector constructed by the cabinet door opening angle, the cabinet door opening time, the internal reference light intensity, and the external reference light intensity, and the target label is defined as the actual power loss corresponding to the simulated state vector; S6-2. Input the loss sample training set into a general neural network model, and generate the loss prediction model after supervised learning iterative training.

7. The method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet according to claim 6, characterized in that: Get the loss sample training set, including: S6-1-1, preset the cabinet door opening angle D1, cabinet door opening time T1; S6-1-2. Under the simulation state of maintaining the external reference light intensity unchanged, open the cabinet door so that the cabinet door opening angle is D1, the cabinet door opening time is T1, and measure the actual power loss SE1 of the simulation state; S6-1-3, construct the external reference light intensity, the cabinet door opening angle D1, the cabinet door opening time T1 and the rated internal reference light intensity as the simulation state vector, and define the actual power loss SE1 as the target label of the simulation state vector; S6-1-4, gradually increase the cabinet door opening angle and cabinet door opening time in an arithmetic progression, and repeat S6-1-2 to generate K different simulation state vectors and corresponding target labels; S6-1-5, combining the generated simulation state vector and the target label to obtain the loss sample; S6-1-6. Arrange the simulation state vectors and the corresponding target labels in a gradually increasing order to construct the loss sample training set.

8. The method for processing light-sensing electrical signals for energy saving in a refrigerator wine cabinet according to claim 6, characterized in that: The loss sample training set is input into a general neural network model, and after supervised learning iterative training, the loss prediction model is generated, including: S6-2-1. Extract the first batch of loss samples and use them as input to the general neural network model; S6-2-2, according to the loss samples of the first batch, perform forward propagation on them, and output the predicted value of the power loss of the first batch in the output layer; S6-3-3, taking the mean square error between the predicted value of the first batch of power loss and the actual power loss of the first batch as the target loss of the first batch; S6-4-4. Calculate the gradient of the target loss of the first batch relative to the current model parameters; S6-4-5, updating the model parameters according to the calculated gradient using the gradient descent method to generate a real-time parameter model; S6-4-6, extract the next batch of loss samples, use them as the model input of the real-time parameter model, and repeat S6-2-2 to S6-2-5 until the output target loss is minimized; S6-2-7. When the target loss is minimized, the real-time parameter model is defined as the loss prediction model.

Images