Air source unit control method and device

Abstract

The present application discloses a control method and device for an air source unit, which monitors and collects indoor temperature and outdoor temperature in real time through a temperature sensor, and introduces a data processing and analysis algorithm at the back end to perform a time series collaborative correlation analysis of the indoor temperature and outdoor temperature, so as to capture the time series characteristics of the indoor and outdoor temperature difference, and automatically adjust the operating parameters of the air source unit according to the changes in the indoor and outdoor temperatures. In this way, automatic adjustment of the operating parameters of the air source unit is achieved, thereby improving the refrigeration efficiency and energy-saving performance.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to an air source unit control method and device. Background Art

[0002] Air source units are commonly used for heating, cooling and hot water supply. They adjust indoor temperature by absorbing or releasing heat from outdoor air, and can achieve ultra-low temperature cooling effect. The operating frequency of air source units is an important factor affecting their cooling performance and energy consumption, so they need to be reasonably adjusted according to changes in indoor and outdoor temperatures.

[0003] However, existing air source unit control methods usually adopt fixed operating frequencies or frequency adjustment strategies based on experience. These methods cannot automatically adjust according to changes in indoor and outdoor temperatures, cannot fully utilize the potential of air source units, and cannot adapt to complex environmental changes.

[0004] Therefore, an optimized air source unit control scheme is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an air source unit control method and device, which monitors and collects indoor temperature and outdoor temperature in real time through a temperature sensor, and introduces a data processing and analysis algorithm at the back end to perform a time series collaborative correlation analysis of the indoor temperature and outdoor temperature, so as to capture the time series characteristics of the indoor and outdoor temperature difference, and automatically adjust the operating parameters of the air source unit according to the changes in indoor and outdoor temperatures. In this way, the automatic adjustment of the operating parameters of the air source unit is achieved, thereby improving the refrigeration efficiency and energy saving performance.

[0006] According to one aspect of the present application, there is provided an air source unit control method, which includes:

[0007] Obtaining a time series of indoor temperature and a time series of outdoor temperature collected by a temperature sensor;

[0008] Sequence segmentation is performed on the time series of the indoor temperature and the time series of the outdoor temperature based on the same time scale to obtain a sequence of a local time series of the indoor temperature and a sequence of a local time series of the outdoor temperature;

[0009] Calculating the positional difference between each group of corresponding local time series of indoor temperature and local time series of outdoor temperature in the sequence of local time series of indoor temperature and the sequence of local time series of outdoor temperature to obtain a sequence of local time series of indoor and outdoor temperature difference;

[0010] Arranging each of the indoor and outdoor temperature difference local time series in the sequence of indoor and outdoor temperature difference local time series into a sequence of indoor and outdoor temperature difference local time input vectors according to the time dimension;

[0011] Performing feature extraction on the sequence of the indoor and outdoor temperature difference local time input vectors by an indoor and outdoor temperature difference time series feature extractor based on a deep neural network model to obtain a sequence of contextual indoor and outdoor temperature difference local time feature vectors;

[0012] Using a prototype feature extraction network to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors to obtain an indoor and outdoor temperature difference prototype time series feature vector as an indoor and outdoor temperature difference prototype time series feature;

[0013] Based on the prototype time series characteristics of the indoor and outdoor temperature difference, determine whether the operating frequency value at the current time point should be increased, decreased or maintained.

[0014] According to another aspect of the present application, there is provided an air source unit control device, comprising:

[0015] A data acquisition module, used to acquire a time series of indoor temperature and a time series of outdoor temperature collected by a temperature sensor;

[0016] A sequence segmentation module, used for segmenting the time series of the indoor temperature and the time series of the outdoor temperature based on the same time scale to obtain a sequence of a local time series of the indoor temperature and a sequence of a local time series of the outdoor temperature;

[0017] An indoor-outdoor temperature difference calculation module, used for calculating the position difference between each corresponding group of the local time series of indoor temperature and the local time series of outdoor temperature in the sequence of the local time series of indoor temperature and the sequence of the local time series of outdoor temperature to obtain a sequence of the local time series of indoor-outdoor temperature difference;

[0018] An arrangement module, used for arranging each of the indoor and outdoor temperature difference local time series in the sequence of the indoor and outdoor temperature difference local time series into a sequence of indoor and outdoor temperature difference local time input vectors according to the time dimension;

[0019] A contextual indoor and outdoor temperature difference local time feature extraction module is used to extract features from the sequence of the indoor and outdoor temperature difference local time input vectors through an indoor and outdoor temperature difference time series feature extractor based on a deep neural network model to obtain a sequence of contextual indoor and outdoor temperature difference local time feature vectors;

[0020] An indoor and outdoor temperature difference prototype time series feature extraction module is used to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors using a prototype feature extraction network to obtain an indoor and outdoor temperature difference prototype time series feature vector as an indoor and outdoor temperature difference prototype time series feature;

[0021] The control result generation module is used to determine whether the operating frequency value at the current time point should be increased, decreased or maintained based on the prototype time series characteristics of the indoor and outdoor temperature difference.

[0022] Compared with the prior art, the present application provides an air source unit control method and device, which monitors and collects indoor temperature and outdoor temperature in real time through temperature sensors, and introduces data processing and analysis algorithms at the back end to perform time series collaborative correlation analysis of the indoor temperature and outdoor temperature, so as to capture the time series characteristics of the indoor and outdoor temperature difference, and automatically adjust the operating parameters of the air source unit according to the changes in indoor and outdoor temperatures. In this way, automatic adjustment of the operating parameters of the air source unit is achieved, thereby improving the cooling efficiency and energy-saving performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0024] Figure 1 is a flow chart of an air source unit control method according to an embodiment of the present application;

[0025] Figure 2 A system architecture diagram of an air source unit control method according to an embodiment of the present application;

[0026] Figure 3 is a flow chart of the training phase of the air source unit control method according to an embodiment of the present application;

[0027] Figure 4 4 is a block diagram of an air source unit control device according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0029] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0030] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0031] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0032] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0033] Existing air source unit control methods usually adopt fixed operating frequencies or frequency adjustment strategies based on experience. These methods cannot automatically adjust according to changes in indoor and outdoor temperatures, cannot fully utilize the potential of air source units, and cannot adapt to complex environmental changes. Therefore, an optimized air source unit control solution is desired.

[0034] In the technical solution of the present application, a method for controlling an air source unit is proposed. Figure 1 Flow chart of an air source unit control method according to an embodiment of the present application. Figure 2 : is a system architecture diagram of the air source unit control method according to an embodiment of the present application. Figure 1 and Figure 2As shown, the air source unit control method according to the embodiment of the present application includes the following steps: S1, acquiring the time series of indoor temperature and the time series of outdoor temperature collected by the temperature sensor; S2, segmenting the time series of indoor temperature and the time series of outdoor temperature based on the same time scale to obtain a sequence of local time series of indoor temperature and a sequence of local time series of outdoor temperature; S3, calculating the position difference between the local time series of indoor temperature and the local time series of outdoor temperature in each group of the sequence of local time series of indoor temperature and the sequence of local time series of outdoor temperature to obtain a sequence of local time series of indoor and outdoor temperature difference; S4, dividing the indoor and outdoor temperature into the local time series of outdoor temperature and the indoor and outdoor temperature. Each indoor and outdoor temperature difference local time series in the sequence of temperature difference local time series is arranged according to the time dimension as a sequence of indoor and outdoor temperature difference local time input vectors; S5, extracting features from the sequence of indoor and outdoor temperature difference local time input vectors by an indoor and outdoor temperature difference time series feature extractor based on a deep neural network model to obtain a sequence of context indoor and outdoor temperature difference local time feature vectors; S6, processing the sequence of context indoor and outdoor temperature difference local time feature vectors using a prototype feature extraction network to obtain an indoor and outdoor temperature difference prototype time series feature vector as an indoor and outdoor temperature difference prototype time series feature; S7, determining whether the operating frequency value at the current time point should be increased, decreased or maintained based on the indoor and outdoor temperature difference prototype time series feature.

[0035] In particular, the S1 obtains the time series of indoor temperature and outdoor temperature collected by the temperature sensor. It is worth mentioning that the temperature sensor is a device used to measure the temperature of an environment or an object. It can convert the temperature into an electrical signal to facilitate the measurement, monitoring and control of the temperature.

[0036] In particular, the S2, based on the same time scale, performs sequence segmentation on the time series of the indoor temperature and the time series of the outdoor temperature to obtain a sequence of a local time series of the indoor temperature and a sequence of a local time series of the outdoor temperature. Considering that the indoor temperature and the outdoor temperature have a dynamic change law and volatility of the time series in the time dimension, that is, the change of temperature is usually nonlinear, and there may be a periodic or seasonal pattern. Therefore, in order to capture more detailed features of the temperature time series change, so as to better analyze the temperature change trend and pattern, in the technical solution of the present application, based on the same time scale, the time series of the indoor temperature and the time series of the outdoor temperature are sequence segmented to obtain a sequence of a local time series of the indoor temperature and a sequence of a local time series of the outdoor temperature. By segmenting the time series of the indoor temperature and the outdoor temperature into local time series, the entire time series can be decomposed into multiple small time periods, and it can also be ensured that the time points corresponding to the local time series of the indoor temperature and the outdoor temperature are consistent, so that the change of the indoor and outdoor temperature difference can be more accurately analyzed, and then the optimal operation parameter adjustment strategy can be determined.

[0037] In particular, the S3 calculates the position difference between each group of the local time series of the indoor temperature and the local time series of the outdoor temperature in the sequence of the local time series of the indoor temperature and the sequence of the local time series of the outdoor temperature to obtain the sequence of the local time series of the indoor and outdoor temperature difference. It should be understood that since the operation frequency control of the air source unit is based on the temperature difference between indoor and outdoor, in order to capture the change of the indoor and outdoor temperature difference for further analysis and control, in the technical solution of the present application, the position difference between each group of the local time series of the indoor temperature and the local time series of the outdoor temperature in the sequence of the local time series of the indoor temperature and the sequence of the local time series of the outdoor temperature is calculated to obtain the sequence of the local time series of the indoor and outdoor temperature difference. It should be understood that the indoor and outdoor temperature difference refers to the difference between the indoor temperature and the outdoor temperature. By calculating the position difference between the local time series of the indoor temperature and the local time series of the outdoor temperature, it is possible to provide information about the time series change trend and fluctuation of the indoor and outdoor temperature difference, which is very important for determining the adjustment strategy of the operating parameters.

[0038] In particular, in S4, each of the indoor and outdoor temperature difference local time series in the sequence of the indoor and outdoor temperature difference local time series is arranged according to the time dimension into a sequence of indoor and outdoor temperature difference local time input vectors. That is, in the technical solution of the present application, in order to be able to perform a more detailed and comprehensive analysis of the temporal variation pattern and trend of the indoor and outdoor temperature difference, so as to more accurately control the operating frequency, in the technical solution of the present application, each of the indoor and outdoor temperature difference local time series in the sequence of the indoor and outdoor temperature difference local time series is further arranged according to the time dimension into a sequence of indoor and outdoor temperature difference local time input vectors.

[0039] In particular, the S5 performs feature extraction on the sequence of the local time input vectors of the indoor and outdoor temperature difference through the indoor and outdoor temperature difference time series feature extractor based on the deep neural network model to obtain a sequence of contextual indoor and outdoor temperature difference local time feature vectors. In particular, the indoor and outdoor temperature difference time series feature extractor based on the deep neural network model is an indoor and outdoor temperature difference time series feature extractor based on the Transformer layer. Specifically, the sequence of the local time input vectors of the indoor and outdoor temperature difference is passed through the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer to extract the local time series contextual associated feature information of the indoor and outdoor temperature difference in each local time period, thereby obtaining a sequence of contextual indoor and outdoor temperature difference local time feature vectors. More specifically, the sequence of the indoor and outdoor temperature difference local time input vectors is arranged in one dimension to obtain a global indoor and outdoor temperature difference feature vector; the product of the global indoor and outdoor temperature difference feature vector and the transposed vector of each indoor and outdoor temperature difference local time input vector in the sequence of the indoor and outdoor temperature difference local time input vector is calculated to obtain a plurality of self-attention association matrices; a standardization unit is used to standardize each of the plurality of self-attention association matrices to obtain a plurality of standardized self-attention association matrices; an attention calculation unit is used to pass each of the plurality of standardized self-attention association matrices through a Softmax classification function to obtain a plurality of probability values; an attention application unit is used to weight each of the plurality of local time input vectors in the sequence of the indoor and outdoor temperature difference local time input vectors using each of the plurality of probability values ​​as a weight to obtain a sequence of the contextual indoor and outdoor temperature difference local time feature vectors.

[0040] The Transformer layer is a neural network layer widely used in natural language processing and other sequence processing tasks. The core idea of ​​the Transformer layer is to use the self-attention mechanism to capture the correlation between different positions in the input sequence. Compared with traditional recurrent neural networks and convolutional neural networks, the Transformer layer can process all positions in the sequence in parallel, thereby improving computational efficiency.

[0041] In particular, the S6 uses a prototype feature extraction network to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors to obtain the indoor and outdoor temperature difference prototype time series feature vector as the indoor and outdoor temperature difference prototype time series feature. It should be understood that since each contextual indoor and outdoor temperature difference local time feature vector in the sequence of the contextual indoor and outdoor temperature difference local time feature vectors respectively represents the local time series context features of the indoor and outdoor temperature difference, in particular, considering that in the actual indoor and outdoor temperature monitoring to control the corresponding operating frequency, the indoor and outdoor temperature differences in different local time periods will show certain time series change rules and trends in specific seasons and specific regions, but occasionally there will be special cases that are affected by external environmental factors and cause the time series change characteristics of the indoor and outdoor temperature difference to be relatively different, which cannot reflect the essential characteristics of the indoor and outdoor temperature difference time series characteristics, and also have a small contribution to the control of the operating frequency, and it is necessary to reduce the influence of the indoor and outdoor temperature difference time series characteristics that deviate from the essential characteristics on the prototype calculation. Based on this, in the technical solution of the present application, the prototype feature extraction network is further used to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors to obtain the indoor and outdoor temperature difference prototype time series feature vector. It should be understood that, through the processing of the prototype feature extraction network, the local temporal context features of the indoor and outdoor temperature difference of each local segment in the sequence of the contextual indoor and outdoor temperature difference local time feature vector can be subjected to weighted fusion analysis based on the prototype essential contribution, so as to map the temporal contextual association features of each local indoor and outdoor temperature difference to a feature space with more distinguishing and expressive capabilities, thereby capturing the essential characterization characteristics of the temporal changes of the indoor and outdoor temperature difference, which contains important information about the temperature difference changes, such as the amplitude, frequency, and trend of the temperature difference. This can improve the temporal feature expression and discrimination capabilities of the indoor and outdoor temperature difference, thereby more accurately realizing the adaptive control of the operating parameters of the air source unit. Specifically, the prototype feature extraction network is used to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors with the following prototype formula to obtain the indoor and outdoor temperature difference prototype time series feature vector; wherein, the prototype formula is:

[0042]

[0043]

[0044] Among them, v i and v j are the i-th and j-th contextual indoor and outdoor temperature difference local time feature vectors in the sequence of the contextual indoor and outdoor temperature difference local time feature vectors, V k is the sequence of the local time feature vectors of the contextual indoor and outdoor temperature difference, ||·||1 is the vector norm, M is the number of vectors in the sequence of the local time feature vectors of the contextual indoor and outdoor temperature difference - 1, Di is the eigenvalue of each position in the semantic feature vector of indoor temperature difference fluctuation, N s is the length of the indoor temperature difference fluctuation semantic feature vector, V is the indoor and outdoor temperature difference prototype time series feature vector, and exp(·) is an exponential operation.

[0045] In particular, the S7 determines whether the operating frequency value at the current time point should be increased, decreased or maintained based on the prototype time series characteristics of the indoor and outdoor temperature difference. That is, in the technical solution of the present application, the prototype time series feature vector of the indoor and outdoor temperature difference is passed through a compressor operating frequency controller based on a classifier to obtain an operating frequency control instruction, and the operating frequency control instruction is used to indicate that the operating frequency value at the current time point should be increased, decreased or maintained. In other words, the prototype time series essential feature information of the indoor and outdoor temperature difference is used for classification processing to automatically control the compressor operating parameters, such as increasing, decreasing or maintaining the operating frequency of the compressor. In this way, the operating parameters of the air source unit can be automatically adjusted according to the changes in indoor and outdoor temperatures, and the automatic adjustment of the operating parameters of the air source unit is realized, thereby improving the refrigeration efficiency and energy-saving performance. Specifically, the prototype time series feature vector of the indoor and outdoor temperature difference is fully connected and encoded using multiple fully connected layers of the classifier to obtain an encoded classification feature vector; and the encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the classification result.

[0046] A classifier is a machine learning model or algorithm that is used to classify input data into different categories or labels. Classifiers are part of supervised learning and perform classification tasks by learning the mapping relationship from input data to output categories.

[0047] A fully connected layer is a common type of layer in neural networks. In a fully connected layer, each neuron is connected to all neurons in the previous layer, and each connection has a weight. This means that each neuron in the fully connected layer receives inputs from all neurons in the previous layer, and performs a weighted sum of these inputs through the weights, and then passes the result to the next layer.

[0048] The Softmax classification function is a commonly used activation function for multi-classification problems. It converts each element of the input vector into a probability value between 0 and 1, and the sum of these probability values ​​is equal to 1. The Softmax function is often used in the output layer of a neural network, especially for multi-classification problems, because it can map the network output to the probability distribution of each category. During the training process, the output of the Softmax function can be used to calculate the loss function and update the network parameters through the back propagation algorithm. It is worth noting that the output of the Softmax function does not change the relative size relationship between the elements, but only normalizes them. Therefore, the Softmax function does not change the characteristics of the input vector, but only converts it into a probability distribution form.

[0049] It should be understood that before using the above-mentioned neural network model for inference, it is necessary to train the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer, the prototype feature extraction network and the classifier-based compressor operating frequency controller. That is to say, in the air source unit control method of the present application, it also includes a training stage for training the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer, the prototype feature extraction network and the classifier-based compressor operating frequency controller.

[0050] Figure 3 FIG. 1 is a flow chart of the training phase of the air source unit control method according to an embodiment of the present application. Figure 3As shown, the air source unit control method according to the embodiment of the present application includes: a training stage, including: S110, obtaining training data, the training data including a time series of a training indoor temperature and a time series of a training outdoor temperature collected by a temperature sensor; S120, performing sequence segmentation on the time series of the training indoor temperature and the time series of the training outdoor temperature based on the same time scale to obtain a sequence of local time series of the training indoor temperature and a sequence of local time series of the training outdoor temperature; S130, calculating the position difference between each group of corresponding local time series of the training indoor temperature and the local time series of the training outdoor temperature in the sequence of local time series of the training indoor temperature and the sequence of local time series of the training outdoor temperature to obtain a sequence of local time series of training indoor and outdoor temperature differences; S140, arranging each local time series of the training indoor and outdoor temperature difference in the sequence of local time series of the training indoor and outdoor temperature difference according to the time dimension as a sequence of local time series of training indoor and outdoor temperature differences. A sequence of input vectors; S150, performing feature extraction on the sequence of training indoor and outdoor temperature difference local time input vectors through the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer to obtain a sequence of training context indoor and outdoor temperature difference local time feature vectors; S160, using a prototype feature extraction network to process the sequence of training context indoor and outdoor temperature difference local time feature vectors to obtain a training indoor and outdoor temperature difference prototype time series feature vector; S170, optimizing the training indoor and outdoor temperature difference prototype time series feature vector to obtain an optimized training indoor and outdoor temperature difference prototype time series feature vector; S180, passing the training indoor and outdoor temperature difference prototype time series feature vector through a classifier-based compressor operating frequency controller to obtain a classification loss function value; S190, training the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer, the prototype feature extraction network and the classifier-based compressor operating frequency controller based on the classification loss function value.

[0051] Among them, the training indoor and outdoor temperature difference prototype time series feature vector is passed through a classifier-based compressor operating frequency controller to obtain a classification loss function value, including: using the classifier to process the training indoor and outdoor temperature difference prototype time series feature vector to obtain a training classification result; and calculating the cross entropy loss function value between the training classification result and the true value of the operating frequency value at the current time point that should be increased, decreased or maintained as the classification loss function value.

[0052] In the technical solution of the present application, each training context indoor and outdoor temperature difference local time feature vector in the sequence of the training context indoor and outdoor temperature difference local time feature vector expresses the cross-time domain context association feature between the local time domain distribution of the training indoor and outdoor temperature difference value. In this way, considering the uneven distribution of source data time series between each training indoor and outdoor temperature difference local time series in the sequence of the training indoor and outdoor temperature difference local time series, after the extraction of the indoor and outdoor temperature difference time domain context association feature, there will also be significant imbalance and inconsistency between the feature distributions of each of the training context indoor and outdoor temperature difference local time feature vectors. Therefore, when the sequence of the training context indoor and outdoor temperature difference local time feature vectors is processed using the prototype feature extraction network, considering that the context indoor and outdoor temperature difference local time feature vectors between local time domains have significant imbalance and inconsistency in feature distribution, the significance of the feature distribution information of the context semantic feature distribution between the time domains of the indoor and outdoor temperature difference of each training context indoor and outdoor temperature difference local time feature vector based on its local time domain will be affected, making it difficult to stably focus on the significant local distribution of features when the training indoor and outdoor temperature difference prototype time series feature vector is classified by the classifier, thereby affecting the training speed. Based on this, the applicant of the present application optimizes the training indoor and outdoor temperature difference prototype time series feature vector each time the training indoor and outdoor temperature difference prototype time series feature vector is iterated through the classification regression by the classifier, expressed as:

[0053]

[0054] Where V is the time series feature vector of the indoor and outdoor temperature difference prototype for training, and are the squares of the 1-norm and 2-norm of the training indoor and outdoor temperature difference prototype time series feature vector, respectively, L is the length of the training indoor and outdoor temperature difference prototype time series feature vector, and ω is a weight hyperparameter, and log represents a logarithm operation with base 2. Specifically, by geometrically registering the high-dimensional feature manifold shape based on the scale and structural parameters of the training indoor and outdoor temperature difference prototype time series feature vector V, it is possible to focus on the features with rich feature semantic information in the feature set composed of the eigenvalues ​​of the training indoor and outdoor temperature difference prototype time series feature vector V, that is, the distinguishable stable interest features based on the local context information representation dissimilarity when the classifier is classified, thereby realizing the significant labeling of the feature information of the training indoor and outdoor temperature difference prototype time series feature vector V in the classification process, and improving the training speed of the classifier. In this way, the operating parameters of the air source unit can be automatically adjusted according to the changes in indoor and outdoor temperatures, thereby improving the cooling efficiency and energy-saving performance.

[0055] In summary, the air source unit control method according to the embodiment of the present application is explained, which monitors and collects the indoor temperature and outdoor temperature in real time through temperature sensors, and introduces data processing and analysis algorithms at the back end to perform time series collaborative correlation analysis of the indoor temperature and outdoor temperature, so as to capture the time series characteristics of the indoor and outdoor temperature difference, and automatically adjust the operating parameters of the air source unit according to the changes in indoor and outdoor temperatures. In this way, automatic adjustment of the operating parameters of the air source unit is achieved, thereby improving the cooling efficiency and energy-saving performance.

[0056] Furthermore, an air source unit control device is also provided.

[0057] Figure 4 FIG. 1 is a block diagram of an air source unit control device according to an embodiment of the present application. Figure 4 As shown, according to the air source unit control device 300 of the embodiment of the present application, it includes: a data acquisition module 310, which is used to obtain the time series of indoor temperature and the time series of outdoor temperature collected by the temperature sensor; a sequence segmentation module 320, which is used to segment the time series of the indoor temperature and the time series of the outdoor temperature based on the same time scale to obtain a sequence of local time series of indoor temperature and a sequence of local time series of outdoor temperature; an indoor and outdoor temperature difference calculation module 330, which is used to calculate the position difference between the local time series of indoor temperature and the local time series of outdoor temperature in each group corresponding to the sequence of local time series of indoor temperature and the sequence of local time series of outdoor temperature to obtain a sequence of local time series of indoor and outdoor temperature difference; an arrangement module 340, which is used to arrange the local time series of indoor and outdoor temperature difference into a sequence of local time series of indoor and outdoor temperature difference. Each local time series of indoor and outdoor temperature difference in the sequence of sequences is arranged according to the time dimension as a sequence of local time input vectors of indoor and outdoor temperature difference; a contextual indoor and outdoor temperature difference local time feature extraction module 350 is used to extract features from the sequence of local time input vectors of indoor and outdoor temperature difference through an indoor and outdoor temperature difference time series feature extractor based on a deep neural network model to obtain a sequence of contextual indoor and outdoor temperature difference local time feature vectors; an indoor and outdoor temperature difference prototype time series feature extraction module 360 ​​is used to process the sequence of contextual indoor and outdoor temperature difference local time feature vectors using a prototype feature extraction network to obtain an indoor and outdoor temperature difference prototype time series feature vector as an indoor and outdoor temperature difference prototype time series feature; a control result generation module 370 is used to determine whether the operating frequency value at the current time point should be increased, decreased or maintained based on the indoor and outdoor temperature difference prototype time series feature.

[0058] As described above, the air source unit control device 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with an air source unit control algorithm. In one possible implementation, the air source unit control device 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the air source unit control device 300 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the air source unit control device 300 can also be one of the many hardware modules of the wireless terminal.

[0059] Alternatively, in another example, the air source unit control device 300 and the wireless terminal may also be separate devices, and the air source unit control device 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0060] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for controlling an air source unit, characterized in that: include: Obtaining a time series of indoor temperature and a time series of outdoor temperature collected by a temperature sensor; Sequence segmentation is performed on the time series of the indoor temperature and the time series of the outdoor temperature based on the same time scale to obtain a sequence of a local time series of the indoor temperature and a sequence of a local time series of the outdoor temperature; Calculating the positional difference between each group of corresponding local time series of indoor temperature and local time series of outdoor temperature in the sequence of local time series of indoor temperature and the sequence of local time series of outdoor temperature to obtain a sequence of local time series of indoor and outdoor temperature difference; Arranging each of the indoor and outdoor temperature difference local time series in the sequence of indoor and outdoor temperature difference local time series into a sequence of indoor and outdoor temperature difference local time input vectors according to the time dimension; Performing feature extraction on the sequence of the indoor and outdoor temperature difference local time input vectors by an indoor and outdoor temperature difference time series feature extractor based on a deep neural network model to obtain a sequence of contextual indoor and outdoor temperature difference local time feature vectors; Using a prototype feature extraction network to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors to obtain an indoor and outdoor temperature difference prototype time series feature vector as an indoor and outdoor temperature difference prototype time series feature; Based on the time series characteristics of the indoor and outdoor temperature difference prototype, determine whether the operating frequency value at the current time point should be increased, decreased or maintained; The method further comprises a training step: for training the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer, the prototype feature extraction network and the compressor operation frequency controller based on the classifier; The training step comprises: Acquire training data, wherein the training data includes a time series of training indoor temperature and a time series of training outdoor temperature collected by a temperature sensor; Sequence segmentation is performed on the time series of the training indoor temperature and the time series of the training outdoor temperature based on the same time scale to obtain a sequence of local time series of the training indoor temperature and a sequence of local time series of the training outdoor temperature; Calculating the positional difference between each group of corresponding local time series of training indoor temperature and local time series of training outdoor temperature in the sequence of local time series of training indoor temperature and the sequence of local time series of training outdoor temperature to obtain a sequence of local time series of training indoor and outdoor temperature difference; Arranging the individual training indoor and outdoor temperature difference local time series in the sequence of training indoor and outdoor temperature difference local time series according to the time dimension into a sequence of training indoor and outdoor temperature difference local time input vectors; The sequence of local time input vectors of the training indoor and outdoor temperature difference is subjected to feature extraction by a Transformer-based indoor and outdoor temperature difference time series feature extractor to obtain a sequence of local time feature vectors of the training context indoor and outdoor temperature difference; Processing the sequence of the training context indoor and outdoor temperature difference local time feature vectors using a prototype feature extraction network to obtain a training indoor and outdoor temperature difference prototype time series feature vector; Optimizing the training indoor and outdoor temperature difference prototype time series feature vector to obtain an optimized training indoor and outdoor temperature difference prototype time series feature vector; The training indoor and outdoor temperature difference prototype time series feature vector is passed through a classifier-based compressor operation frequency controller to obtain a classification loss function value; Based on the classification loss function value, the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer, the prototype feature extraction network and the classifier-based compressor operation frequency controller are trained; Each time the training indoor and outdoor temperature difference prototype time series feature vector is iterated through classification regression by the classifier, the training indoor and outdoor temperature difference prototype time series feature vector is optimized, which is expressed as: Where V is the time series feature vector of the indoor and outdoor temperature difference prototype for training, and are the squares of the 1-norm and 2-norm of the prototype time series feature vector of the indoor and outdoor temperature difference in the training, respectively, L is the length of the prototype time series feature vector of the indoor and outdoor temperature difference in the training, and ω is a weight hyperparameter, and log represents a logarithmic operation with base 2; The method comprises: using the prototype feature extraction network to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors to obtain the indoor and outdoor temperature difference prototype time series feature vector as the indoor and outdoor temperature difference prototype time series feature, comprising: using the prototype feature extraction network to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors with the following prototype formula to obtain the indoor and outdoor temperature difference prototype time series feature vector; Among them, the prototype formula is: Among them, v i and v j are the i-th and j-th contextual indoor and outdoor temperature difference local time feature vectors in the sequence of the contextual indoor and outdoor temperature difference local time feature vectors, V k is the sequence of the local time feature vectors of the contextual indoor and outdoor temperature difference, ||·||1 is the vector norm, M is the number of vectors in the sequence of the local time feature vectors of the contextual indoor and outdoor temperature difference - 1, D i is the eigenvalue of each position in the semantic feature vector of indoor temperature difference fluctuation, N s is the length of the indoor temperature difference fluctuation semantic feature vector, V is the indoor and outdoor temperature difference prototype time series feature vector, and exp(·) is an exponential operation.

2. The air source unit control method according to claim 1, characterized in that: Based on the prototype timing characteristics of the indoor and outdoor temperature difference, determine whether the operating frequency value at the current time point should be increased, decreased or maintained, including: passing the prototype timing characteristic vector of the indoor and outdoor temperature difference through a classifier-based compressor operating frequency controller to obtain an operating frequency control instruction, and the operating frequency control instruction is used to indicate that the operating frequency value at the current time point should be increased, decreased or maintained.

3. The air source unit control method according to claim 2, characterized in that: The training indoor and outdoor temperature difference prototype time series feature vector is passed through a classifier-based compressor operation frequency controller to obtain a classification loss function value, including: Using the classifier to process the training indoor and outdoor temperature difference prototype time series feature vector to obtain a training classification result: and The cross entropy loss function value between the training classification result and the true value of the operating frequency value at the current time point that should be increased, decreased, or maintained is calculated as the classification loss function value.

4. An air source unit control device, characterized in that: include: A data acquisition module, used to obtain a time series of indoor temperature and a time series of outdoor temperature collected by a temperature sensor; A sequence segmentation module, used for segmenting the time series of the indoor temperature and the time series of the outdoor temperature based on the same time scale to obtain a sequence of a local time series of the indoor temperature and a sequence of a local time series of the outdoor temperature; An indoor-outdoor temperature difference calculation module, used for calculating the position difference between each corresponding group of the local time series of indoor temperature and the local time series of outdoor temperature in the sequence of the local time series of indoor temperature and the sequence of the local time series of outdoor temperature to obtain a sequence of the local time series of indoor-outdoor temperature difference; An arrangement module, used for arranging each of the indoor and outdoor temperature difference local time series in the sequence of indoor and outdoor temperature difference local time series into a sequence of indoor and outdoor temperature difference local time input vectors according to the time dimension; A contextual indoor and outdoor temperature difference local time feature extraction module is used to extract features from the sequence of the indoor and outdoor temperature difference local time input vectors through an indoor and outdoor temperature difference time series feature extractor based on a deep neural network model to obtain a sequence of contextual indoor and outdoor temperature difference local time feature vectors; An indoor and outdoor temperature difference prototype time series feature extraction module is used to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors using a prototype feature extraction network to obtain an indoor and outdoor temperature difference prototype time series feature vector as an indoor and outdoor temperature difference prototype time series feature; A control result generation module is used to determine whether the operating frequency value at the current time point should be increased, decreased or maintained based on the time series characteristics of the indoor and outdoor temperature difference prototype; The device also includes a training module: used for training the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer, the prototype feature extraction network and the compressor operation frequency controller based on the classifier; The training module is used to: Acquire training data, wherein the training data includes a time series of training indoor temperature and a time series of training outdoor temperature collected by a temperature sensor; Sequence segmentation is performed on the time series of the training indoor temperature and the time series of the training outdoor temperature based on the same time scale to obtain a sequence of local time series of the training indoor temperature and a sequence of local time series of the training outdoor temperature; Calculating the positional difference between each group of corresponding local time series of training indoor temperature and local time series of training outdoor temperature in the sequence of local time series of training indoor temperature and the sequence of local time series of training outdoor temperature to obtain a sequence of local time series of training indoor and outdoor temperature difference; Arranging the individual training indoor and outdoor temperature difference local time series in the sequence of training indoor and outdoor temperature difference local time series according to the time dimension into a sequence of training indoor and outdoor temperature difference local time input vectors; The sequence of local time input vectors of the training indoor and outdoor temperature difference is subjected to feature extraction by a Transformer-based indoor and outdoor temperature difference time series feature extractor to obtain a sequence of local time feature vectors of the training context indoor and outdoor temperature difference; Processing the sequence of the training context indoor and outdoor temperature difference local time feature vectors using a prototype feature extraction network to obtain a training indoor and outdoor temperature difference prototype time series feature vector; Optimizing the training indoor and outdoor temperature difference prototype time series feature vector to obtain an optimized training indoor and outdoor temperature difference prototype time series feature vector; The training indoor and outdoor temperature difference prototype time series feature vector is passed through a classifier-based compressor operation frequency controller to obtain a classification loss function value; Based on the classification loss function value, the indoor and outdoor temperature difference time series feature extractor based on the Transformer layer, the prototype feature extraction network and the classifier-based compressor operation frequency controller are trained; Each time the training indoor and outdoor temperature difference prototype time series feature vector is iterated through classification regression by the classifier, the training indoor and outdoor temperature difference prototype time series feature vector is optimized, which is expressed as: Where V is the time series feature vector of the indoor and outdoor temperature difference prototype for training, and are the squares of the 1-norm and 2-norm of the prototype time series feature vector of the indoor and outdoor temperature difference in the training, respectively, L is the length of the prototype time series feature vector of the indoor and outdoor temperature difference in the training, and ω is a weight hyperparameter, and log represents a logarithmic operation with base 2; The indoor and outdoor temperature difference prototype time series feature extraction module is used to: use the prototype feature extraction network to process the sequence of the contextual indoor and outdoor temperature difference local time feature vectors with the following prototype formula to obtain the indoor and outdoor temperature difference prototype time series feature vector; Among them, the prototype formula is: Among them, v i and v j are the i-th and j-th contextual indoor and outdoor temperature difference local time feature vectors in the sequence of the contextual indoor and outdoor temperature difference local time feature vectors, V k is the sequence of the local time feature vectors of the contextual indoor and outdoor temperature difference, ||·||1 is the vector norm, M is the number of vectors in the sequence of the local time feature vectors of the contextual indoor and outdoor temperature difference - 1, D i is the eigenvalue of each position in the semantic feature vector of indoor temperature difference fluctuation, N s is the length of the indoor temperature difference fluctuation semantic feature vector, V is the indoor and outdoor temperature difference prototype time series feature vector, and exp(·) is an exponential operation.

Images