Compressor dynamic protection control method and electronic starter controller

Abstract

The invention provides a dynamic protection control method for a compressor and an electronic starter controller, and the method comprises the steps: carrying out the scene labeling of the operation data of the compressor, dividing various load scenes, and setting a basic critical value offset coefficient; parameters of the compressor are collected in real time, the real-time load intensity is calculated through a weighting algorithm, and a buffer area is determined; the method comprises the following steps: acquiring data of a compressor, converting the data into a time sequence feature sequence, performing normalization processing, then inputting the time sequence feature sequence into a pre-trained neural network model of a double-branch network structure, and outputting a trend level and a PID parameter interval; according to the trend level and the PID parameter interval, determining a basic PID parameter, and introducing a risk deviation coefficient to determine a final PID parameter; and smooth transition amplitude limiting is set for the final output quantity of the PID parameters, the compressor is operated under the smooth transition amplitude limiting, and the problems of time delay and curve fluctuation of traditional critical value control are solved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of compressor dynamic protection control, and particularly relates to a compressor dynamic protection control method and an electronic starting controller. BACKGROUND

[0002] A compressor is a driven fluid machine that raises low-pressure gas to high-pressure gas, and it absorbs low-temperature and low-pressure refrigerant gas from a suction pipe, compresses the refrigerant gas, and discharges high-temperature and high-pressure refrigerant gas to a discharge pipe to provide power for a refrigeration cycle.

[0003] In the traditional compressor over-temperature or over-current protection, a critical value is set, and if the temperature or current exceeds the critical value, the control of the compressor is started, but this control has a certain control delay, which still causes a short over-temperature or over-current; in addition, the control curve of the entire compressor fluctuates. SUMMARY

[0004] Therefore, the present application provides a compressor dynamic protection control method and an electronic starting controller, which aims to solve the delay and curve fluctuation problems of the traditional critical value control, and realize the early intervention and smooth regulation of the compressor.

[0005] The first aspect of the present application provides a compressor dynamic protection control method, which performs scene tagging on the operation data of the compressor, divides various load scenes, and sets a basic critical value offset coefficient for each load scene; Real-time parameters of the compressor are collected, and a real-time load intensity is calculated through a weighting algorithm, wherein the parameters at least include an operating current, a winding temperature, and a refrigerant pressure; A buffer interval is determined according to a traditional critical value, the load intensity, and the basic critical value offset coefficient; Data of the compressor are acquired and converted into a time sequence feature sequence, data in the time sequence feature sequence are normalized based on the buffer interval, and then input into a pre-trained neural network model of a double-branch network structure, to output a trend level and a PID parameter interval, wherein the data of the compressor include the operating current, the winding temperature, the refrigerant pressure, an outdoor environment temperature, an indoor set temperature, a fan speed, a winding temperature change rate within a preset time, and an operating current change rate within a preset time, the neural network model of the double-branch network structure includes a trend prediction branch and a PID parameter generation branch, the trend prediction branch is a network of an LSTM layer + Attention, the PID parameter generation branch is a network of a full connection layer + residual network, and in addition, when the data fall into the buffer interval, the Attention mechanism gives higher weights to the corresponding data; According to the trend level and the PID parameter interval, a basic PID parameter is determined, and a risk deviation coefficient is introduced, and according to the basic PID parameter and the risk deviation coefficient, a final PID parameter is determined; The output quantity of the final PID parameter is set to a smooth transition limiting range, and the compressor is operated under the smooth transition limiting range.

[0006] Further, in the step of collecting the parameters of the compressor in real time and calculating the real-time load intensity through a weighting algorithm, the weights corresponding to the parameters of the compressor are obtained through historical fault data training.

[0007] Further, the step of obtaining the weights corresponding to the parameters of the compressor through historical fault data training comprises: obtaining historical operation data of the compressor, and extracting current features, temperature features, pressure features and labels from the historical operation data; After data cleaning and enhancement of the extracted current features, temperature features, pressure features and labels, a logistic regression is used as a weight learning model, and the weights are optimized through gradient descent in the training process to determine the final weights.

[0008] Further, in the step of determining the buffer interval according to the traditional critical value, the load intensity and the basic critical value offset coefficient, the calculation formula of the buffer interval is: upper limit of the buffer interval = traditional critical value × (1-basic critical value offset coefficient × S); lower limit of the buffer interval = traditional critical value × (1-basic critical value offset coefficient × S-0.1); wherein, S is the load intensity.

[0009] Further, the step of obtaining the data of the compressor and converting it into a time sequence feature sequence, normalizing the data in the time sequence feature sequence based on the buffer interval, and then inputting it into a pre-trained neural network model with a double-branch network structure to output the trend level and the PID parameter interval comprises: obtaining the data of the compressor and converting it into a time sequence feature sequence, normalizing the data in the time sequence feature sequence based on the buffer interval, and then inputting it into an LSTM layer to output a global time sequence feature vector; determining a risk-sensitive time sequence feature according to the global time sequence feature vector and the buffer interval; outputting a trend level according to the risk-sensitive time sequence feature; concatenating the trend level and the working condition features as the input of a full connection layer, wherein a residual network structure is added to prevent parameter mapping overfitting; The result of the residual network is input into an output layer, and PID parameters are mapped to a target interval according to Sigmoid activation.

[0010] Further, the step of determining a basic PID parameter according to the trend level and the PID parameter interval and introducing a risk deviation coefficient, and determining a final PID parameter according to the basic PID parameter and the risk deviation coefficient comprises: determining a basic PID parameter according to the trend level and the PID parameter interval; quantifying the risk emergency degree of the current basic PID parameter corresponding to the position in the buffer area according to the risk deviation coefficient, dynamically amplifying / reducing the regulation strength of the basic PID parameter, and determining a final PID parameter.

[0011] Further, in the step of quantifying the risk emergency degree of the current basic PID parameter corresponding to the position in the buffer area according to the risk deviation coefficient, the calculation formula of the risk emergency degree is: risk deviation coefficient = (upper limit of buffer interval - actual value) / (upper limit of buffer interval - lower limit of buffer interval); corrected deviation = traditional deviation x (2 - risk deviation coefficient).

[0012] The second aspect of the embodiment of the application provides an electronic starting controller for implementing the compressor dynamic protection control method. The division module is configured to perform scene labeling on the operation data of the compressor, divide various load scenes, and set a basic critical value offset coefficient for each load scene. The calculation module is configured to collect parameters of the compressor in real time, and calculate a real-time load intensity by using a weighting algorithm, wherein the parameters at least include an operating current, a winding temperature, and a refrigerant pressure. The first determination module is configured to determine a buffer interval according to a traditional critical value, the load intensity, and the basic critical value offset coefficient. The input module is used for acquiring data of the compressor and converting the data into a time sequence feature sequence, performing normalization processing on data in the time sequence feature sequence based on the buffer interval, and then inputting the neural network model of the pre-trained double-branch network structure to output a trend level and a PID parameter interval, wherein the data of the compressor includes the operating current, the winding temperature, the refrigerant pressure, the outdoor environment temperature, the indoor set temperature, the fan speed, the winding temperature change rate within a preset time, and the operating current change rate within a preset time, wherein the neural network model of the double-branch network structure includes a trend prediction branch and a PID parameter generation branch, the trend prediction branch is an LSTM layer + Attention network, and the PID parameter generation branch is a full connection layer + residual network. The second determination module is used for determining a basic PID parameter according to the trend level and the PID parameter interval, introducing a risk deviation coefficient, determining a final PID parameter according to the basic PID parameter and the risk deviation coefficient. The setting module is used for setting a smooth transition limiting amplitude for an output of the final PID parameter, and running the compressor under the smooth transition limiting amplitude.

[0013] The compressor dynamic protection control method and the electronic starting controller provided in the embodiment of the application divide various load scenes by performing scene labeling on the operating data of the compressor, set a basic critical value offset coefficient for each load scene, collect parameters of the compressor in real time, calculate a real-time load intensity through a weighting algorithm, the parameters at least include the operating current, the winding temperature, and the refrigerant pressure, determine a buffer interval according to a traditional critical value, the load intensity, and the basic critical value offset coefficient, acquire data of the compressor, convert the data into a time sequence feature sequence, perform normalization processing on data in the time sequence feature sequence based on the buffer interval, then input the neural network model of the pre-trained double-branch network structure to output a trend level and a PID parameter interval, determine a basic PID parameter according to the trend level and the PID parameter interval, introduce a risk deviation coefficient, determine a final PID parameter according to the basic PID parameter and the risk deviation coefficient, set a smooth transition limiting amplitude for an output of the final PID parameter, and run the compressor under the smooth transition limiting amplitude, specifically, the traditional fixed critical value thinking is broken through, the scene adaptation of the buffer interval is realized through the load intensity calculation, the problem of inaccurate pre-intervention timing is solved, the trend prediction is combined with the PID parameter generation, the PID is changed from passive response to active prediction, and the regulation and control fluctuation is reduced from the root. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 The implementation flowchart of the compressor dynamic protection control method provided for the first embodiment of the application is shown in FIG. 1. Figure 2 Figure 1 is a block diagram of an electronic control device according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] For the purpose of promoting an understanding of the principles of the application, reference will now be made to the embodiments illustrated in the drawings. It is expressly understood that the drawings are presented for the purpose of illustration and description only and are not intended as a definition of the limits of the application. As illustrated in the drawings:

[0016] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. It is also possible that the use of the same terms can be different depending on the context in which they are used. For example, the term "connected" can mean directly connected or connected through intervening components.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting. The use of the terms "and / or" includes a combination of one or more of the associated listed items.

[0018] Embodiment One According to the embodiments of the present application, a compressor dynamic protection control method is provided. It is to be understood that the steps shown in the flowcharts of the drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0019] In this embodiment one, a compressor dynamic protection control method is provided. Please refer to Figure 1 , Figure 1 Figure 1 is a block diagram of a compressor dynamic protection control method according to an embodiment of the present application, which specifically includes steps S01 to S06.

[0020] Step S01, the operation data of the compressor is scene-labeled, various load scenarios are divided, and a basic critical value offset coefficient is set for each load scenario.

[0021] It should be noted that the prior art uses a fixed critical value for protection control, which cannot adapt to compressor load fluctuations (such as summer refrigeration peak, winter heating start, voltage fluctuation, etc.), resulting in a buffer interval that is either too narrow (not effective for pre-intervention) or too wide (wasting energy). Therefore, in the embodiments of the present application, the scenario adaptation of the buffer interval is realized by load intensity calculation, solving the problem of inaccurate pre-intervention timing.

[0022] Specifically, the running data of the compressor throughout its life cycle is first labeled by scene, and six types of core load scenes are divided, including a light load scene (such as low load operation in spring and autumn, indoor and outdoor temperature difference <5℃), a conventional load scene (such as stable operation in summer during the day, temperature difference 5-10℃), a heavy load scene (such as high-temperature start in summer after noon, temperature difference >10℃), a start-up impact scene (current instantaneous peak value within 10 minutes after the compressor is powered on), a voltage fluctuation scene (when the grid voltage fluctuates ±10%), and an abnormal approach scene (such as slow temperature rise due to condenser blockage, approaching the traditional critical value). Each scene corresponds to a set of basic critical value offset coefficients (such as a light load scene offset coefficient of 0.2 and a heavy load scene offset coefficient of 0.3), avoiding the disadvantages of a single buffer interval adapting to all scenes.

[0023] Step S02, real-time acquisition of parameters of the compressor, and calculation of real-time load intensity by a weighted algorithm, the parameters at least including operating current, winding temperature and refrigerant pressure.

[0024] Specifically, the parameters of the compressor are collected in real time, including operating current I, winding temperature T and refrigerant pressure P, and the load intensity S is calculated by weighted summation, wherein the weights corresponding to the operating current I, winding temperature T and refrigerant pressure P are obtained by training historical fault data.

[0025] More specifically, the historical running data of the compressor is obtained, and current characteristics, temperature characteristics, pressure characteristics and labels are extracted from the historical running data. In the embodiments of the present application, the historical running data of the compressor at least contains 3000 complete running cycles, of which at least 500 over-temperature / over-current fault cases are included, covering different working conditions (such as high temperature in summer, low temperature in winter, voltage fluctuation, etc.) and equipment states (such as new machine, running for 1 year, running for 3 years), it can be understood that the current characteristic refers to the average current value within 1 minute before the fault occurs, the temperature characteristic refers to the average winding temperature within 1 minute before the fault occurs, the pressure characteristic refers to the average refrigerant high pressure within 1 minute before the fault occurs, and the label includes 1 and 0, 1 indicating that the running state corresponding to the group of characteristics eventually leads to over-temperature / over-current fault, and 0 indicating that the running state corresponding to the group of characteristics does not cause fault (normal operation); After data cleaning and enhancement of the extracted current features, temperature features, pressure features, and labels, a logistic regression is used as a weight learning model. In the training process, the weights are optimized by gradient descent to determine the final weights. The output of the logistic regression is the probability of failure, which can directly reflect the correlation strength between the features and the failure risk. The model parameters (coefficients) can be directly used as the weights of each feature. In addition, a cross-entropy loss function is used to measure the prediction error of the model, and the goal is to minimize the loss.

[0026] Step S03, determining the buffer interval according to the traditional critical value, the load intensity, and the basic critical value offset coefficient.

[0027] Specifically, the calculation formula of the buffer interval is: Upper limit of the buffer interval = traditional critical value x (1-basic critical value offset coefficient x S); Lower limit of the buffer interval = traditional critical value x (1-basic critical value offset coefficient x S-0.1); Wherein, S is the load intensity.

[0028] For example, in the heavy load scenario, S=0.8, the heavy load scenario offset coefficient=0.3, the traditional critical temperature is 80℃, and the buffer interval is 80x(1-0.30.8)~80x(1-0.3x0.8-0.1)=60.8℃~52.8℃. The pre-intervention is 20℃ in advance to avoid temperature "top".

[0029] Step S04, obtaining the data of the compressor and converting it into a time sequence feature sequence. The data in the time sequence feature sequence is normalized based on the buffer interval, and then input into the pre-trained neural network model with a double-branch network structure to output the trend level and the PID parameter interval.

[0030] In the embodiment of the present application, the neural network model with a double-branch network structure is used to identify the parameter variation trend in the buffer interval in advance and provide predictive parameters for PID control to avoid PID control fluctuations caused by lag response. Specifically, the data of the compressor includes the operating current, the winding temperature, the refrigerant pressure, the outdoor environment temperature, the indoor set temperature, the fan speed, the winding temperature change rate within a preset time, the operating current change rate within a preset time, the operating current, the winding temperature, and the refrigerant pressure are core risk features, the outdoor environment temperature, the indoor set temperature, and the fan speed are working condition features, and the winding temperature change rate within a preset time and the operating current change rate within a preset time are historical trend features. The neural network model with a double-branch network structure includes a trend prediction branch and a PID parameter generation branch. The trend prediction branch is an LSTM layer + Attention network, and the PID parameter generation branch is a fully connected layer + residual network. In addition, when the data falls into the buffer interval, the Attention mechanism gives higher weights to the corresponding data. Understandably, among the features input into the neural network model, only when the parameters enter the buffer interval, these features have high risk warning value. For example, when the winding temperature is far below the lower limit of the buffer interval, a slight fluctuation in the winding temperature change rate within a preset time is meaningless for system safety, and the neural network will automatically reduce its weight. When the winding temperature enters the buffer interval, the same winding temperature change rate within a preset time will be given a very high weight (strengthened through the Attention mechanism), ensuring that the model focuses on the scene that really needs intervention.

[0031] It should be noted that the data of the compressor is obtained and converted into a time sequence feature sequence. The data in the time sequence feature sequence is normalized based on the buffer interval, and then input into the LSTM layer to output a global time sequence feature vector. The LSTM layer includes a first LSTM layer and a second LSTM layer. The first LSTM layer includes 32 neurons for learning local time sequence patterns and uses a tanh activation function to retain time sequence direction information. The second LSTM layer includes 16 neurons for learning global trend features and adds Dropout (0.2) to suppress overfitting. Understandably, the neural network model normalizes the input features based on the buffer interval rather than the traditional critical value. For example, the normalization formula for the winding temperature feature is: T_norm=(current winding temperature-buffer interval lower limit) / (buffer interval upper limit-buffer interval lower limit); So that the feature value accurately reflects the risk distance from the upper limit of the buffer interval in the [0, 1] interval, rather than the absolute distance from the traditional critical value, making the model more sensitive to subtle changes within the buffer interval; Based on the global time-series feature vector and the buffer, risk-sensitive time-series features are determined. Specifically, firstly, a risk query vector is constructed, consisting of the upper limit of the buffer, the current parameter value, and historical fault-related features. Then, the similarity between the global time-series feature vector and the risk query vector is calculated to obtain the attention weight of each feature. High-risk features (features falling into the buffer) are forcibly given a higher weight (multiplied by a factor of 1.5) to ensure that the model prioritizes critical risk signals. Finally, the global time-series feature vector and the attention weight are multiplied element by element to obtain the risk-sensitive time-series features. Based on the risk-sensitive time series characteristics, a trend level is output. In this embodiment of the invention, there are 5 trend levels, including sudden drop, gradual drop, stable, gradual rise, and sudden rise. The trend level and working condition features are concatenated and used as the input to the fully connected layer. A residual network structure is added to prevent overfitting of the parameter mapping. It should be noted that the residual network structure includes three residual blocks in series. Each residual block contains two fully connected layers and a skip connection. The first layer has 64 neurons (ReLU activation) and the second layer has 32 neurons (ReLU activation). The skip connection directly superimposes the input to the output (solving the gradient vanishing problem in deep networks). The results of the residual network are input into the output layer. Based on the Sigmoid activation, the PID parameters are mapped to the target interval. In this embodiment of the invention, the Sigmoid interval mapping is as follows: Kp = 0.1 + 0.9 × Sigmoid(Kp_raw) (mapped to [0.1, 1.0]); Ki=0.01+0.09×Sigmoid(Ki_raw) (mapped to [0.01,0.1]); Kd = 0.001 + 0.009 × Sigmoid(Kd_raw) (mapped to [0.001, 0.01]); Kp is the proportional coefficient, Ki is the integral coefficient, Kd is the differential coefficient, Kp_raw is the original proportional coefficient, Ki_raw is the original integral coefficient, and Kd_raw is the original differential coefficient. The original proportional coefficient, original integral coefficient, and original differential coefficient are the results of the residual network.

[0032] In addition, the trend prediction branch of the neural network will judge in real time whether the current buffer zone is reasonable. For example, when the model predicts that "the winding temperature will rise sharply at a rate of 0.8℃ / s", and the current buffer zone only reserves a buffer space of 5℃ (theoretically, 5℃ / 0.8℃ / s=6.25s response time), but the actual PID control requires at least 8s to suppress the temperature rise, the model will indirectly promote the temporary expansion of the buffer zone (such as adding a 2℃ buffer) by outputting a high-risk trend level, so as to avoid the control failure caused by the preset range being too narrow.

[0033] Step S05, according to the trend level and the PID parameter interval, determine the basic PID parameter, and introduce the risk deviation coefficient, according to the basic PID parameter and the risk deviation coefficient, determine the final PID parameter.

[0034] Specifically, according to the trend level and the PID parameter interval, determine the basic PID parameter, exemplary, if the trend is "rapid rise": give larger Kp (enhance proportional control strength, quickly suppress parameter rise), smaller Ki (avoid integral accumulation leading to overshoot), moderate Kd (weaken differential oscillation); If the trend is "slow rise": give moderate Kp, larger Ki (slowly offset static deviation, avoid temperature fluctuation), smaller Kd (reduce the sensitivity of differential to small changes); If the trend is "stable": give smaller Kp, moderate Ki, smaller Kd (maintain stability, avoid over-regulation); According to the risk deviation coefficient, the risk emergency degree of the current basic PID parameter corresponding to the position in the buffer area is quantified, the control strength of the basic PID parameter is dynamically amplified / reduced, and the final PID parameter is determined, it can be understood that the trend type is also slow rise, and the risk degree of 1℃ from the upper limit of the buffer interval and 5℃ from the upper limit is different, so this step is to accurately match the parameter with the risk distance; The calculation formula of quantifying the risk emergency degree is: Risk deviation coefficient=(buffer interval upper limit-actual value) / (buffer interval upper limit-buffer interval lower limit); Corrected deviation=traditional deviation x(2-risk deviation coefficient).

[0035] Exemplary, the actual winding temperature is 60℃, the buffer interval is 52.8℃~60.8℃, the risk deviation=(60.8-60) / (60.8-52.8)=0.1, the corrected deviation=(60.8-60)(2-0.1)=0.81.9=1.52, which is larger than the traditional deviation 0.8, and the PID control strength is stronger, which avoids the winding temperature breaking through the upper limit of the buffer interval.

[0036] Step S06, set the smooth transition limit of the output of the final PID parameter, and run the compressor under the smooth transition limit.

[0037] Specifically, the single control quantity change amplitude is less than or equal to the maximum allowed change, the control quantity change rate is less than or equal to the maximum change rate, and the PID output is processed by a first-order low-pass filter, which further weakens the instantaneous fluctuation of the control quantity, and ensures the smooth transition of the compressor running state.

[0038] In summary, the compressor dynamic protection control method in the above embodiments of the present invention involves: labeling the compressor's operating data to classify various load scenarios and setting basic critical value offset coefficients for each load scenario; collecting compressor parameters in real time and calculating the real-time load intensity using a weighted algorithm, with parameters including at least operating current, winding temperature, and refrigerant pressure; determining a buffer zone based on traditional critical values, load intensity, and basic critical value offset coefficients; acquiring compressor data and converting it into a time-series feature sequence; normalizing the data in the time-series feature sequence using the buffer zone as a reference; and then inputting the data into a pre-trained dual-branch network. In the neural network model of the structure, the output trend level and PID parameter range are output. Based on the trend level and PID parameter range, the basic PID parameters are determined, and a risk deviation coefficient is introduced. Based on the basic PID parameters and the risk deviation coefficient, the final PID parameters are determined. A smooth transition limit is set for the output of the final PID parameters, and the compressor runs under the smooth transition limit. Specifically, it breaks through the traditional fixed critical value thinking and realizes the scenario adaptation between buffer zones through load intensity calculation, solving the problem of inaccurate pre-intervention timing. At the same time, it combines trend prediction with PID parameter generation, so that PID changes from passive response to active prediction, reducing control fluctuations from the root.

[0039] Example 2 Please see Figure 2 , Figure 2 This is a structural block diagram of an electronic starter 200 provided in Embodiment 2 of the present invention. The electronic starter 200 is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0040] Specifically, the electronic starter 200 includes: a division module 21, a calculation module 22, a first determination module 23, an input module 24, a second determination module 25, and a setting module 26, wherein: The segmentation module 21 is used to tag the compressor's operating data into various load scenarios and set basic critical value offset coefficients for each load scenario. The calculation module 22 is used to collect the parameters of the compressor in real time and calculate the real-time load intensity through a weighted algorithm. The parameters include at least the operating current, winding temperature and refrigerant pressure. The weights corresponding to the collected compressor parameters are obtained by training with historical fault data. The first determining module 23 is used to determine the buffer zone based on the traditional critical value, the load intensity, and the basic critical value offset coefficient. The calculation formula for the buffer zone is: Upper limit of buffer interval = traditional critical value * (1 - basic critical value offset coefficient * S); Lower limit of buffer interval = traditional critical value * (1 - basic critical value offset coefficient * S - 0.1); Wherein, S is the load intensity; The input module 24 is configured to acquire data of the compressor, convert the data into a time sequence feature sequence, normalize data in the time sequence feature sequence based on the buffer interval, and then input the data into a pre-trained neural network model of a double-branch network structure to output a trend level and a PID parameter interval, wherein the data of the compressor includes the operating current, the winding temperature, the refrigerant pressure, the outdoor environment temperature, the indoor set temperature, the fan speed, the winding temperature change rate within a preset time, and the operating current change rate within a preset time, the neural network model of the double-branch network structure includes a trend prediction branch and a PID parameter generation branch, the trend prediction branch is an LSTM layer + Attention network, and the PID parameter generation branch is a fully connected layer + residual network. The second determination module 25 is configured to determine a basic PID parameter based on the trend level and the PID parameter interval, introduce a risk deviation coefficient, and determine a final PID parameter based on the basic PID parameter and the risk deviation coefficient. The setting module 26 is configured to set a smooth transition limit for an output of the final PID parameter, and run the compressor under the smooth transition limit.

[0041] Further, in some optional embodiments of the present application, the calculation module 22 includes: An acquisition unit is configured to acquire historical operation data of the compressor, and extract current features, temperature features, pressure features, and labels from the historical operation data. A weight determination unit is configured to, after data cleaning and enhancement of the extracted current features, temperature features, pressure features, and labels, use a logistic regression as a weight learning model, optimize the weight through gradient descent in a training process, and determine a final weight.

[0042] Further, in some optional embodiments of the present application, the input module 24 includes: A normalization processing unit is configured to acquire data of the compressor, convert the data into a time sequence feature sequence, normalize data in the time sequence feature sequence based on the buffer interval, and then input the data into an LSTM layer to output a global time sequence feature vector. A first determination unit is configured to determine a risk-sensitive time sequence feature based on the global time sequence feature vector and the buffer interval. an output unit configured to output a trend level according to the risk-sensitive timing feature; a concatenation unit configured to concatenate the trend level and a working condition feature as an input of a full connection layer, wherein a residual network structure is added to prevent parameter mapping overfitting; a mapping unit configured to input a result of the residual network into an output layer, and map PID parameters to a target interval according to Sigmoid activation.

[0043] Further, in some optional embodiments of the present application, the second determination module 25 comprises: a second determination unit configured to determine a basic PID parameter according to the trend level and the PID parameter interval; a regulation unit configured to determine a final PID parameter by dynamically amplifying / reducing a regulation strength of the basic PID parameter according to a risk deviation coefficient, and quantifying a risk emergency degree of the basic PID parameter corresponding to a position in a buffer region, and a calculation formula of the quantified risk emergency degree is: a risk deviation coefficient=(an upper limit of the buffer interval- an actual value) / (an upper limit of the buffer interval- a lower limit of the buffer interval); a corrected deviation= a traditional deviation x (2- the risk deviation coefficient).

[0044] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily mean the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0045] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the patent scope of the present application. It should be noted that, for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which are all within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A dynamic protection control method for a compressor, characterized in that, The method includes: The compressor's operating data is tagged with scenarios to classify various load scenarios, and a basic critical value offset coefficient is set for each type of load scenario. The compressor parameters are collected in real time, and the real-time load intensity is calculated using a weighted algorithm. The parameters include at least the operating current, winding temperature, and refrigerant pressure. The buffer zone is determined based on the traditional critical value, the load intensity, and the basic critical value offset coefficient; The compressor data is acquired and converted into a time-series feature sequence. The data in the time-series feature sequence is normalized based on the buffer space. Then, it is input into a pre-trained neural network model with a dual-branch network structure. The output is the trend level and PID parameter range. The compressor data includes the operating current, the winding temperature, the refrigerant pressure, the outdoor ambient temperature, the indoor set temperature, the fan speed, the rate of change of the winding temperature within a preset time, and the rate of change of the operating current within a preset time. The neural network model with a dual-branch network structure includes a trend prediction branch and a PID parameter generation branch. The trend prediction branch is an LSTM layer + Attention network, and the PID parameter generation branch is a fully connected layer + residual network network. In addition, when data falls into the buffer space, the Attention mechanism assigns higher weights to the corresponding data. Based on the trend level and the PID parameter range, the basic PID parameters are determined, and a risk deviation coefficient is introduced. Based on the basic PID parameters and the risk deviation coefficient, the final PID parameters are determined. Set a smooth transition limit for the final PID parameter output, and run the compressor under the smooth transition limit.

2. The compressor dynamic protection control method according to claim 1, characterized in that, In the step of real-time acquisition of compressor parameters and calculation of real-time load intensity using a weighted algorithm, the weights corresponding to the acquired compressor parameters are obtained through training with historical fault data.

3. The compressor dynamic protection control method according to claim 2, characterized in that, The steps involved in obtaining the weights corresponding to the parameters of the compressor through training with historical fault data include: Obtain historical operating data of the compressor, and extract current characteristics, temperature characteristics, pressure characteristics, and tags from the historical operating data; After data cleaning and enhancement of the extracted current features, temperature features, pressure features, and labels, logistic regression is used as the weight learning model. During the training process, gradient descent is used to optimize the weights and determine the final weights.

4. The compressor dynamic protection control method according to claim 3, characterized in that, In the step of determining the buffer zone based on the traditional critical value, the load intensity, and the basic critical value offset coefficient, the calculation formula for the buffer zone is as follows: The upper limit between buffer zones = traditional critical value × (1 - basic critical value offset coefficient × S); The lower limit between buffer zones = traditional critical value × (1 - basic critical value offset coefficient × S - 0.1); Wherein, S represents the load intensity.

5. The compressor dynamic protection control method according to claim 4, characterized in that, The steps of acquiring compressor data, converting it into a time-series feature sequence, normalizing the data in the time-series feature sequence based on the buffer interval, and then inputting it into a pre-trained neural network model with a dual-branch network structure to output trend levels and PID parameter ranges include: The compressor data is acquired and converted into a time-series feature sequence. The data in the time-series feature sequence is normalized based on the buffer space, and then input into the LSTM layer to output a global time-series feature vector. Based on the global time series feature vector and the buffer, risk-sensitive time series features are determined; Based on the aforementioned risk-sensitive time-series characteristics, output the trend level; The trend level is concatenated with the working condition features and used as the input to the fully connected layer, in which a residual network structure is added to prevent overfitting of parameter mapping. The results of the residual network are input into the output layer, and the PID parameters are mapped to the target range based on the Sigmoid activation.

6. The compressor dynamic protection control method according to claim 5, characterized in that, The steps of determining the basic PID parameters based on the trend level and the PID parameter range, introducing a risk deviation coefficient, and determining the final PID parameters based on the basic PID parameters and the risk deviation coefficient include: Based on the trend level and the PID parameter range, determine the basic PID parameters; Based on the risk deviation coefficient, the risk urgency of the current basic PID parameters corresponding to the position within the buffer zone is quantified, and the control intensity of the basic PID parameters is dynamically amplified / reduced to determine the final PID parameters.

7. The compressor dynamic protection control method according to claim 6, characterized in that, In the step of quantifying the risk urgency of the current basic PID parameters corresponding to the position within the buffer zone based on the risk deviation coefficient, the formula for calculating the risk urgency is as follows: Risk deviation coefficient = (upper limit between buffer zones - actual value) / (upper limit between buffer zones - lower limit between buffer zones); Corrected deviation = Traditional deviation × (2 - Risk deviation coefficient).

8. An electronic starter, characterized in that, For implementing the compressor dynamic protection control method as described in any one of claims 1-7, the electronic starter includes: The segmentation module is used to tag the compressor's operating data into different scenarios, classify them into various load scenarios, and set basic critical value offset coefficients for each type of load scenario. The calculation module is used to collect compressor parameters in real time and calculate the real-time load intensity through a weighted algorithm. The parameters include at least the operating current, winding temperature and refrigerant pressure. The first determining module is used to determine the buffer zone based on the traditional critical value, the load intensity, and the basic critical value offset coefficient. The input module is used to acquire compressor data and convert it into a time-series feature sequence. The data in the time-series feature sequence is normalized based on the buffer space. Then, it is input into a pre-trained neural network model with a dual-branch network structure, which outputs the trend level and PID parameter range. The compressor data includes the operating current, the winding temperature, the refrigerant pressure, the outdoor ambient temperature, the indoor set temperature, the fan speed, the rate of change of the winding temperature within a preset time, and the rate of change of the operating current within a preset time. The neural network model with a dual-branch network structure includes a trend prediction branch and a PID parameter generation branch. The trend prediction branch is an LSTM layer + Attention network, and the PID parameter generation branch is a fully connected layer + residual network network. In addition, when data falls into the buffer space, the Attention mechanism assigns higher weights to the corresponding data. The second determining module is used to determine the basic PID parameters based on the trend level and the PID parameter range, introduce a risk deviation coefficient, and determine the final PID parameters based on the basic PID parameters and the risk deviation coefficient. The setting module is used to set a smooth transition limit for the output of the final PID parameters, and to run the compressor under the smooth transition limit.

Images