Cold capacity determination method and device, electronic equipment and readable storage medium

Abstract

The invention relates to the technical field of cold quantity prediction, and provides a cold quantity determination method and device, electronic equipment and a readable storage medium. The method comprises the steps that at least one piece of cooling capacity characteristic data corresponding to a to-be-processed object is input into a trained cooling capacity determination model for embedding processing, and embedded vectors corresponding to the to-be-processed object are obtained; performing gating mechanism processing on each embedded vector corresponding to the to-be-processed object to obtain a hidden state vector corresponding to the to-be-processed object; processing the hidden state vector corresponding to the to-be-processed object through a preset activation function to obtain target cooling capacity data corresponding to the to-be-processed object; and the target cooling capacity data corresponding to the to-be-processed object is sent to the target terminal equipment for cooling capacity control, so that the problems of low accuracy and poor generalization ability of a cooling capacity determination model in cooling capacity determination of a big data model due to lack of historical data in a newly-built shopping mall in the prior art are solved, the accuracy is improved, and the user experience is improved. And the applicability of the cold capacity determination model is enhanced.

Description

Technical Field

[0001] The present disclosure relates to the technical field of cooling capacity prediction, and in particular to a cooling capacity determination method, device, electronic device and readable storage medium. Background Art

[0002] The main function of a shopping mall's refrigeration system is to maintain the indoor environment within a suitable temperature range, ensure that customers and goods are in a comfortable state, avoid discomfort and damage to goods caused by high temperatures, promote air circulation, reduce humidity, improve shopping experience, and ensure the air quality and hygiene standards inside the mall. The mall's refrigeration system relies heavily on the determination of cooling capacity. In the mall's cooling capacity prediction, since newly built shopping malls often lack historical data, traditional big data-based prediction models cannot be used directly, which will lead to reduced prediction accuracy and difficulties in energy consumption management.

[0003] It can be seen that the existing technology has problems such as low accuracy in determining the cooling capacity of the big data model and poor generalization ability of the cooling capacity determination model due to the lack of historical data for newly built shopping malls. Summary of the invention

[0004] In view of this, the embodiments of the present disclosure provide a method, device, electronic device and readable storage medium for determining cooling capacity, so as to solve the problems in the prior art of low accuracy and poor generalization ability of cooling capacity determination models in big data models due to lack of historical data of newly built shopping malls.

[0005] According to a first aspect of an embodiment of the present disclosure, a method for determining cooling capacity is provided, comprising: inputting at least one cooling capacity feature data corresponding to an object to be processed into a trained cooling capacity determination model for embedding processing to obtain respective embedding vectors corresponding to the object to be processed; performing gating mechanism processing on respective embedding vectors corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed; processing the hidden state vector corresponding to the object to be processed by a preset activation function to obtain target cooling capacity data corresponding to the object to be processed; and sending the target cooling capacity data corresponding to the object to be processed to a target terminal device for cooling capacity control.

[0006] According to a second aspect of an embodiment of the present disclosure, a device for determining cooling capacity is provided, including: a first processing module, for inputting at least one cooling capacity feature data corresponding to an object to be processed into a trained cooling capacity determination model for embedding processing to obtain respective embedding vectors corresponding to the object to be processed; a second processing module, for performing gating mechanism processing on respective embedding vectors corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed; a third processing module, for processing the hidden state vector corresponding to the object to be processed by a preset activation function to obtain target cooling capacity data corresponding to the object to be processed; and a fourth processing module, for sending the target cooling capacity data corresponding to the object to be processed to a target terminal device for cooling capacity control.

[0007] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0008] According to a fourth aspect of an embodiment of the present disclosure, a readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0009] Compared with the prior art, the disclosed embodiment has the following beneficial effects: at least one cold quantity feature data corresponding to the object to be processed is embedded through the embedding layer in the trained cold quantity determination model to obtain each embedding vector corresponding to the object to be processed, and then each embedding vector corresponding to the object to be processed can be gated through the gated loop unit to obtain a hidden state vector corresponding to the object to be processed, and based on a preset activation function, the hidden state vector corresponding to the object to be processed is activated and transformed, and the hidden state vector corresponding to the object to be processed is converted into target cold quantity data corresponding to the object to be processed that is easy to read and understand, and then the target cold quantity data corresponding to the object to be processed can be sent to the target terminal device for releasing and controlling the cold quantity, thereby improving the accuracy of cold quantity determination and enhancing the versatility of the cold quantity determination model. The data is converted into a vector through the embedding layer, which improves the processing efficiency of the cold quantity determination model, enhances the degree of understanding of the data, improves the scalability of the cold quantity determination model, and ensures the real-time and effectiveness of cold quantity determination. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0011] Figure 1 is a schematic diagram of an application scenario of an embodiment of the present disclosure;

[0012] Figure 2 is a flow chart of a method for determining cooling capacity provided by an embodiment of the present disclosure;

[0013] Figure 3 is a structural schematic diagram of a device for determining cooling capacity provided by an embodiment of the present disclosure;

[0014] Figure 4 It is a structural schematic diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0015] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0016] It should be noted that the user information (including but not limited to terminal device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.

[0017] A method and device for determining cooling capacity according to an embodiment of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0018] Figure 1 Schematic diagram of an application scenario of an embodiment of the present disclosure. The application scenario may include terminal devices 1, 2 and 3, a server 4 and a network 5.

[0019] Terminal devices 1, 2 and 3 can be hardware or software. When terminal devices 1, 2 and 3 are hardware, they can be various electronic devices with display screens and supporting communication with server 4, including but not limited to wall-mounted air conditioners, floor-standing air conditioners, central air conditioners, and refrigeration control panels, etc.; when terminal devices 1, 2 and 3 are software, they can be installed in the above electronic devices. Terminal devices 1, 2 and 3 can be implemented as multiple software or software modules, or as a single software or software module, and the embodiments of the present disclosure are not limited to this. Furthermore, various applications can be installed on terminal devices 1, 2 and 3, such as data processing applications, instant messaging tools, social platform software, search applications, shopping applications, etc.

[0020] The server 4 may be a server that provides various services, for example, a background server that receives a request sent by a terminal device that establishes a communication connection with the server, and the background server may receive and analyze the request sent by the terminal device, and generate a processing result. The server 4 may be a single server, or a server cluster composed of several servers, or a cloud computing service center, which is not limited in the embodiments of the present disclosure.

[0021] It should be noted that the server 4 can be hardware or software. When the server 4 is hardware, it can be various electronic devices that provide various services for the terminal devices 1, 2, and 3. When the server 4 is software, it can be multiple software or software modules that provide various services for the terminal devices 1, 2, and 3, or it can be a single software or software module that provides various services for the terminal devices 1, 2, and 3, and the embodiments of the present disclosure are not limited to this.

[0022] The network 5 can be a wired network connected by coaxial cable, twisted pair and optical fiber, or it can be a wireless network that can interconnect various communication devices without wiring, such as Bluetooth, Near Field Communication (NFC), infrared, etc., which is not limited in the embodiments of the present disclosure.

[0023] The user can establish a communication connection with the server 4 via the network 5 through the terminal devices 1, 2 and 3 to receive or send information, etc. Specifically, the server 4 can embed at least one cold feature data corresponding to the object to be processed through the embedding layer in the trained cold determination model to obtain each embedding vector corresponding to the object to be processed, and then the gated loop unit can be used to perform a gating mechanism on each embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed, and based on a preset activation function, the hidden state vector corresponding to the object to be processed is activated and transformed, and the hidden state vector corresponding to the object to be processed is converted into the target cold data corresponding to the object to be processed that is easy to read and understand, and then the target cold data corresponding to the object to be processed can be sent to the target terminal device to release and control the cold.

[0024] It should be noted that the specific types, quantities and combinations of the terminal devices 1, 2 and 3, the server 4 and the network 5 can be adjusted according to the actual needs of the application scenario, and the embodiments of the present disclosure are not limited to this.

[0025] Figure 2 It is a flow chart of a method for determining cooling capacity provided in an embodiment of the present disclosure. Figure 2 The cooling capacity can be determined by Figure 1 The server executes. Figure 2As shown, the method for determining the cooling capacity includes:

[0026] Step 201: input at least one cooling characteristic data corresponding to the object to be processed into the trained cooling determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed.

[0027] Specifically, at least one cold capacity characteristic data of the object to be processed can be used as input, wherein the object to be processed can be a building, industrial equipment, or a cold chain logistics system, etc., which is not limited here. The cold capacity characteristic data includes but is not limited to ambient temperature, humidity, heat load, and / or equipment operating status, etc., and is input into a cold capacity determination model that has been pre-trained with sample data. The cold capacity determination model can perform nonlinear transformation and embedding processing on at least one cold capacity characteristic data corresponding to the object to be processed through deep learning or machine learning algorithms, and then obtain an embedding vector corresponding to the object to be processed, thereby enhancing the interpretability of the data and the accuracy of the cold capacity determination model, improving the accuracy of cold capacity demand prediction, and thereby improving energy utilization efficiency and stability.

[0028] The object to be processed may be an entity whose cooling demand needs to be evaluated or optimized, including but not limited to buildings, industrial equipment, or cold chain logistics systems.

[0029] The cooling characteristic data corresponding to the object to be processed may be parameters used to characterize the cooling demand or state change of the object to be processed, including but not limited to ambient temperature, humidity, heat load, equipment operating status, and / or historical energy consumption data.

[0030] The cooling capacity determination model can be a prediction model built based on a machine learning or deep learning algorithm. The cooling capacity determination model can be trained by learning the relationship between a large amount of historical data and features. The cooling capacity determination model can be used to determine the cooling capacity demand of the object to be processed.

[0031] The embedded vector corresponding to the object to be processed can be a low-dimensional or high-dimensional vector representation obtained after embedding the cold feature data corresponding to the object to be processed through the cold determination model. The embedded vector corresponding to the object to be processed contains the key information of the cold feature data corresponding to the object to be processed, and deeper features and patterns are extracted through the learning ability of the cold determination model.

[0032] For example, in cold chain logistics, each refrigerated compartment can be an object to be processed, and the temperature, humidity, type and quantity of loaded goods inside the refrigerated compartment can be used as cold capacity characteristic data. The cold capacity characteristic data can be input into the cold capacity determination model, and the embedding vector corresponding to each refrigerated compartment can be obtained after embedding processing.

[0033] Step 202: Perform gating mechanism processing on each embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed.

[0034] Specifically, the embedded vector corresponding to the object to be processed can be processed through a gating mechanism, wherein the gating mechanism can be implemented through a forget gate, an input gate, and an output gate in a long short-term memory network LSTM, or an update gate and a reset gate in a gated recurrent unit GRU, etc., which are not limited here. The gating mechanism can then control the inflow and outflow of information to obtain long-term dependencies in sequence data, and can perform weighted fusion on the input at the current moment and the hidden state at the previous moment to obtain a hidden state vector corresponding to the object to be processed, thereby enhancing the ability of the cooling determination model to process sequence data, improving the adaptability to dynamically changing cooling demands, and improving the accuracy and robustness of the cooling determination model's predictions. Through the gating mechanism, the cooling determination model can more effectively capture the changes in cooling characteristics of the object to be processed at different time points, thereby improving the accuracy of future cooling demand predictions.

[0035] Among them, the hidden state vector corresponding to the object to be processed can be a vector representation obtained after processing the embedded vector through the gating mechanism. The hidden state vector corresponding to the object to be processed can be used to characterize the cooling demand state and context information of the object to be processed at a specific time point. The hidden state vector corresponding to the object to be processed can be obtained through the structure and algorithm of the gating mechanism. The hidden state vector corresponding to the object to be processed contains the embedded vector information at the current time point, and the historical information is integrated through the gating mechanism.

[0036] For example, in the air-conditioning system of an intelligent building, the object to be processed can be each room or area, and the cooling characteristic data can be the temperature, humidity, personnel activity, etc. corresponding to the room or area. The cooling characteristic data can be converted into an embedded vector and then processed through the gating mechanism of the gated cyclic unit to obtain the hidden state vector corresponding to each room or area.

[0037] Step 203: Process the hidden state vector corresponding to the object to be processed by a preset activation function to obtain target cooling data corresponding to the object to be processed.

[0038] Specifically, the hidden state vector corresponding to the object to be processed can be input into a preset activation function. The preset activation function includes but is not limited to Sigmoid, ReLU, or Tanh. The preset activation function can perform a nonlinear transformation on the input vector, and then obtain the target cooling data corresponding to the object to be processed, thereby enhancing the ability of the cooling determination model to handle complex data relationships, improving the accuracy of predicting target cooling data, and improving the generalization ability of the cooling determination model.

[0039] The preset activation function may be a function predefined in the neural network, and the preset activation function may be used to introduce nonlinear factors and map the input of the neuron to the output, so that the neural network can learn and solve nonlinear problems.

[0040] The target cooling capacity data corresponding to the object to be processed may refer to the cooling capacity demand or output value related to the object to be processed predicted or calculated through the cooling capacity determination model under the condition of given cooling capacity characteristic data. The target cooling capacity data corresponding to the object to be processed may be obtained through hidden layer processing of the cooling capacity determination model and nonlinear transformation of the activation function. The target cooling capacity data corresponding to the object to be processed may be used to guide the operation of the equipment, optimize energy use or perform fault diagnosis, etc.

[0041] For example, in an intelligent building management system, the environmental parameters of each area in the building and the operating status of the equipment can be collected as the cooling characteristic data corresponding to the object to be processed, where the environmental parameters include but are not limited to temperature or humidity, etc. The cooling characteristic data can then be processed through the hidden layer of the cooling determination model, and can be nonlinearly transformed through the preset ReLU function to obtain the target cooling data required for each area.

[0042] Step 204: Send the target cooling capacity data corresponding to the object to be processed to the target terminal device for cooling capacity control.

[0043] Specifically, the target cooling data corresponding to the object to be processed can be sent to the designated target terminal device through the communication protocol and communication interface, wherein the communication protocol can be the transmission control protocol, the Internet protocol, or the message queue telemetry transmission, etc., which is not limited here, and the communication interface can be the network interface card or the wireless communication module, etc., which is not limited here, thereby enhancing the degree of automation of the cooling determination method, improving the accuracy and response speed of cooling control, improving energy utilization efficiency and user comfort, and reducing human intervention and errors by sending the target cooling data directly to the terminal device.

[0044] Among them, the target terminal device can be a device that can receive, process and respond to control instructions or data from the server side. The target terminal device includes but is not limited to an intelligent air-conditioning control system, a controller of refrigeration equipment, or an intelligent building management system, etc. The target terminal device can receive and analyze the transmitted target cooling data through a built-in communication module and control system, and then adjust the working state of the target terminal device according to the target cooling data to achieve precise cooling control.

[0045] For example, in a management system of a large data center, the target cooling data required for each area can be calculated through the cooling determination model according to the real-time environmental parameters of the data center and the working status of the server, where the real-time environmental parameters include but are not limited to temperature or humidity. The target cooling data can be sent to the intelligent air-conditioning system in the data center through a high-speed communication network. The intelligent air-conditioning system can adjust the operating parameters of the corresponding refrigeration equipment. The operating parameters can be customs or cooling capacity, etc., which are not limited here.

[0046] According to the technical solution provided by the embodiment of the present disclosure, at least one cold feature data corresponding to the object to be processed is embedded through the embedding layer in the trained cold determination model to obtain each embedding vector corresponding to the object to be processed, and then the gated loop unit can be used to perform gating mechanism processing on each embedding vector corresponding to the object to be processed to obtain the hidden state vector corresponding to the object to be processed, and the hidden state vector corresponding to the object to be processed is activated and transformed based on a preset activation function, and the hidden state vector corresponding to the object to be processed is converted into target cold data corresponding to the object to be processed that is easy to read and understand, and then the target cold data corresponding to the object to be processed can be sent to the target terminal device for releasing and controlling the cold, thereby improving the accuracy of cold determination and enhancing the versatility of the cold determination model. The data is converted into a vector through the embedding layer, which improves the processing efficiency of the cold determination model, enhances the degree of understanding of the data, improves the scalability of the cold determination model, and ensures the real-time and effectiveness of cold determination.

[0047] In some embodiments, before inputting at least one cold capacity characteristic data corresponding to the object to be processed into the trained cold capacity determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed, it also includes: obtaining at least one historical cold capacity related data corresponding to the object to be processed; performing data type check processing on each historical cold capacity related data corresponding to the object to be processed to obtain date data corresponding to the object to be processed and numerical characteristic data corresponding to the object to be processed; performing feature construction processing on the date data corresponding to the object to be processed to obtain time data corresponding to the object to be processed; performing standardization processing on the numerical characteristic data corresponding to the object to be processed to obtain standardized data corresponding to the object to be processed; formatting the time data corresponding to the object to be processed and the standardized data corresponding to the object to be processed to obtain at least one cold capacity characteristic data corresponding to the object to be processed.

[0048] Specifically, at least one historical cooling capacity related data corresponding to the object to be processed can be obtained from a database or a real-time data source. The historical cooling capacity related data includes but is not limited to temperature records, energy consumption data, or equipment status, etc. The data type check can be performed on the historical cooling capacity related data to obtain date data and numerical feature data, wherein the date data includes but is not limited to date or timestamp, etc., and the numerical feature data can be temperature value, humidity value, and / or energy consumption, etc., which are not limited here. The date data can be subjected to feature construction processing, and the date data can be converted into time dimensions such as year, month, day, hour, etc., and it can be determined whether it is a weekend, whether it is a shift holiday, whether it is a holiday, etc. to obtain the corresponding time feature. The numerical feature data can be standardized to eliminate the influence of different dimensions and magnitudes, and then the time data corresponding to the object to be processed and the standardized data corresponding to the object to be processed can be formatted and integrated into the input format required by the cooling capacity determination model, that is, the cooling capacity feature data corresponding to the object to be processed.

[0049] Among them, the historical cooling capacity related data corresponding to the object to be processed can be data records related to the object to be processed in the past period of time. The historical cooling capacity related data corresponding to the object to be processed can be used to characterize the cooling capacity demand or output characteristics of the object to be processed. The historical cooling capacity related data corresponding to the object to be processed can come from sensors, monitoring systems, log files, etc.

[0050] The date data corresponding to the object to be processed can be a data item representing time or date information in the historical cooling-related data. The date data corresponding to the object to be processed can be used to analyze the relationship between cooling demand and time, including but not limited to seasonal changes or day and night differences.

[0051] The numerical characteristic data corresponding to the object to be processed may refer to data items in the historical cooling-related data, which are expressed in numerical form and can quantitatively describe the cooling characteristics of the object to be processed. The numerical characteristic data corresponding to the object to be processed may include temperature, humidity, energy consumption, equipment operating status, etc., which are not limited here.

[0052] The time data corresponding to the object to be processed may be the time dimension information extracted from the date data through feature construction processing. The time data corresponding to the object to be processed may be used as the input of the cooling capacity determination model. The time data corresponding to the object to be processed may include year, month, day, hour, week, etc., and may also include time information such as whether it is a weekend, whether it is a adjusted holiday, whether it is a holiday, etc. The time data corresponding to the object to be processed may also be used to capture the law of changes in cooling demand over time.

[0053] The standardized data corresponding to the object to be processed can be a numerical value after standardization processing. The standardization process can convert each numerical value in the numerical feature data into a dimensionless numerical value with a uniform magnitude. The standardization process eliminates the dimensional differences and magnitude differences between different features.

[0054] For example, in an intelligent refrigeration system, historical cooling capacity-related data such as temperature records, energy consumption data, and equipment status of the target area in the past three years can be obtained, and data type check processing can be performed on the historical cooling capacity-related data to distinguish the historical cooling capacity-related data into date data and numerical feature data. Feature construction processing can be performed on the date data to obtain time data containing time information, and the numerical feature data can be standardized with unified dimensions to obtain dimensionless standardized data with the same order of magnitude. The time data and the standardized data can be formatted to obtain the cooling capacity feature data corresponding to the target area.

[0055] According to the technical solution provided by the embodiment of the present disclosure, at least one historical cold capacity related data corresponding to the object to be processed is obtained from a database or a real-time data source, and a data type check can be performed on the historical cold capacity related data to obtain date data and numerical feature data. Feature construction processing can be performed on the date data to obtain the corresponding time feature, and the numerical feature data can be standardized. Then, the time data corresponding to the object to be processed and the standardized data corresponding to the object to be processed can be formatted and integrated into the input format required by the cold capacity determination model, that is, the cold capacity feature data corresponding to the object to be processed, thereby enhancing the availability and accuracy of the data, improving the understanding and learning ability of the cold capacity determination model for the cold capacity feature data, and improving the accuracy and efficiency of cold capacity prediction or control. Moreover, through data type checking, feature construction, standardization, and formatting, the data is made more standardized and noise and redundancy are reduced.

[0056] In some embodiments, before inputting at least one cold quantity characteristic data corresponding to the object to be processed into the trained cold quantity determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed, it also includes: obtaining a cold quantity training data set corresponding to multiple training objects, each cold quantity training data set contains multiple cold quantity training data and real cold quantity data corresponding to the training object; inputting the cold quantity training data set corresponding to each training object into the cold quantity determination model for embedding processing to obtain an embedding vector set corresponding to each training object, and the embedding vector set corresponding to each training object contains multiple embedding vectors corresponding to the training object; performing gating mechanism processing on the embedding vector set corresponding to each training object to obtain a hidden state vector corresponding to each training object; processing the hidden state vector corresponding to each training object by a preset activation function to obtain the training cold quantity data corresponding to each training object; determining the mean square error loss corresponding to the cold quantity determination model according to the training cold quantity data corresponding to each training object and the real cold quantity data corresponding to each training object; updating the parameters in the cold quantity determination model according to the mean square error loss corresponding to the cold quantity determination model through a loop iteration.

[0057] Specifically, cold volume training data sets can be collected from multiple training objects, each cold volume training data set contains multiple cold volume training data and real cold volume data corresponding to each training object, and then the cold volume training data sets can be respectively input into the cold volume determination model, and the cold volume training data can be converted into an embedding vector set through embedding processing, each embedding vector set contains multiple embedding vectors, and then each embedding vector set can be processed through a gating mechanism to capture the dependency and timing information between data to obtain a hidden state vector, and the hidden state vector can be processed through a preset activation function to obtain training cold volume data, and then the mean square error loss can be calculated based on the difference between the training cold volume data and the real cold volume data as an evaluation indicator of the performance of the cold volume determination model, and the parameters in the cold volume determination model can be adjusted and optimized according to the mean square error loss through a cyclic iteration method to improve the prediction accuracy and generalization ability of the model.

[0058] Among them, the cold capacity training data set corresponding to the training object can be a data set used to train the cold capacity determination model. The cold capacity training data set corresponding to each training object contains the cold capacity training data of each training object under different conditions and the corresponding real cold capacity data. The training data set corresponding to the training object can be obtained through historical records, experimental data, or simulation, etc., which is not limited here.

[0059] The cooling capacity training data may be data used as input for the cooling capacity determination model. The cooling capacity training data may include various types of numerical or category information such as temperature, humidity, energy consumption, and / or equipment status, which are not limited here.

[0060] The actual cooling data corresponding to the training object may be actual cooling data corresponding to the cooling training data, and the actual cooling data may be used to evaluate the accuracy and reliability of the prediction result of the cooling determination model.

[0061] The embedding vector set corresponding to the training object can be a set of vectors obtained by converting the cold training data through embedding processing. Each embedding vector can be a representation of the training object in a specific dimensional space and can be used to capture the characteristics or attributes of the training object.

[0062] The embedding vector corresponding to the training object may be a single vector in the embedding vector set. The embedding vector corresponding to the training object may be used to represent a point or position of the training object in the dimensional space. The embedding vector corresponding to the training object may be obtained through embedding processing.

[0063] The hidden state vector corresponding to the training object can be a vector obtained by processing through a gating mechanism. The hidden state vector corresponding to the training object can be converted from an embedded vector set. The hidden state vector contains the original feature information of the training object and incorporates the dependency and timing information between data.

[0064] The training cooling data corresponding to the training object can be the cooling data obtained after predicting the training object through the cooling determination model. The training cooling data corresponding to the training object can be the prediction result obtained by the cooling determination model based on the training data. The training cooling data corresponding to the training object can be used for comparison and evaluation with the actual cooling data.

[0065] The mean square error loss corresponding to the cooling capacity determination model can be a loss function used to measure the difference between the cooling capacity data predicted by the cooling capacity determination model and the actual cooling capacity data during the training process of the cooling capacity determination model. The smaller the mean square error loss, the closer the prediction result of the cooling capacity determination model is to the actual value and the better the performance.

[0066] For example, in an intelligent building management system, in order to accurately predict the cooling demand of each area in a building, building environment data from multiple historical periods can be obtained as the cooling training data set corresponding to the training object, where the building environment data can be temperature, humidity, and / or personnel flow, etc., which are not limited here. The building environment data set can then be input into the cooling determination model for training, and the key features and information in the data can be extracted through embedded processing and gating mechanisms. During the training process, the cooling determination model can calculate the mean square error loss based on the difference between the predicted cooling data and the actual cooling data, and can optimize the model parameters through a cyclic iteration.

[0067] According to the technical solution provided by the embodiment of the present disclosure, by collecting cold volume training data sets from multiple training objects, each cold volume training data set contains multiple cold volume training data and real cold volume data corresponding to each training object, and then the cold volume training data sets can be respectively input into the cold volume determination model, and the cold volume training data can be converted into an embedding vector set through embedding processing, each embedding vector set contains multiple embedding vectors, and then each embedding vector set can be processed by a gating mechanism to obtain a hidden state vector, and the hidden state vector can be processed by a preset activation function to obtain training cold volume data, and then the mean square error loss can be calculated according to the difference between the training cold volume data and the real cold volume data, and the parameters in the cold volume determination model can be adjusted and optimized according to the mean square error loss through a cyclic iteration method, thereby enhancing the processing capability and learning capability of the cold volume determination model for complex cold volume data, improving the accuracy and stability of the cold volume determination model in predicting cold volume data, and improving the reliability and practicality of the cold volume determination model in practical applications.

[0068] In some embodiments, at least one cold quantity characteristic data corresponding to the object to be processed is input into a trained cold quantity determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed, including: performing data classification processing on at least one cold quantity characteristic data corresponding to the object to be processed to obtain category characteristic data corresponding to the object to be processed, numerical characteristic data corresponding to the object to be processed, and Boolean characteristic data corresponding to the object to be processed; performing vectorized embedding processing on the category characteristic data corresponding to the object to be processed, the numerical characteristic data corresponding to the object to be processed, and the Boolean characteristic data corresponding to the object to be processed, respectively, to obtain each embedding vector corresponding to the object to be processed.

[0069] Specifically, data classification processing can be performed on at least one cooling characteristic data corresponding to the object to be processed, and the cooling characteristic data corresponding to the object to be processed can be divided into category characteristic data, numerical characteristic data, and Boolean characteristic data, wherein the category characteristic data includes but is not limited to equipment type or working mode, the numerical characteristic data includes but is not limited to quantifiable data such as temperature, humidity, and / or energy consumption, and the Boolean characteristic data can be binary data such as switch status or whether it is faulty, which is not limited here. Then, different types of characteristic data can be processed by corresponding vectorized embedding methods, and the category characteristic data can be processed by One-Hot Encoding, the numerical characteristic data can be processed by normalization or standardization methods, and the Boolean characteristic data can be mapped to values ​​of 0 or 1, so as to obtain each embedded vector corresponding to the object to be processed.

[0070] Among them, the category feature data corresponding to the object to be processed can be data describing the category or attribute of the object to be processed. The category feature data corresponding to the object to be processed can be discrete, non-numerical data, including but not limited to device type or working mode, etc. The category feature data corresponding to the object to be processed can be obtained through classification processing, and the category feature data can be used to distinguish different categories of objects to be processed or describe specific attributes.

[0071] The numerical characteristic data corresponding to the object to be processed may be data describing the quantity or measurement relationship of the object to be processed. The numerical characteristic data corresponding to the object to be processed may be continuous and quantifiable. The numerical characteristic data includes but is not limited to temperature, humidity, and / or energy consumption, etc. The numerical characteristic data may be obtained through measurement or calculation and may be used to characterize the physical quantity or state of the object to be processed.

[0072] The Boolean feature data corresponding to the object to be processed can be data that describes whether the state of the object to be processed exists. The Boolean feature data can be binary, that is, there are only two possible results, including but not limited to on / off, yes / no, true / false, etc. The Boolean feature data can be obtained through detection or judgment, and can be used to characterize the state or condition of the object to be processed.

[0073] For example, in an intelligent temperature control system, relevant cooling characteristic data of the room can be collected, which may include the equipment type, current temperature, humidity, and / or number of people in the room, etc., and the equipment type can be used as category characteristic data, the temperature and humidity can be used as numerical characteristic data, and whether there is someone in the room can be used as Boolean characteristic data. These three different types of characteristic data can be vectorized and embedded respectively to obtain corresponding embedding vectors, and the obtained embedding vectors can be input into the trained cooling determination model. The cooling determination model can predict the cooling demand of the room based on the embedded vector, and use this to adjust the operating status of the temperature control equipment.

[0074] According to the technical solution provided by the embodiment of the present disclosure, by performing data classification processing on at least one cold quantity characteristic data corresponding to the object to be processed, the cold quantity characteristic data corresponding to the object to be processed is distinguished into category characteristic data, numerical characteristic data, and Boolean characteristic data, the category characteristic data can be processed by one-hot encoding, the numerical characteristic data can be processed by normalization or standardization method, and the Boolean characteristic data can be mapped to a value of 0 or 1, and then the respective embedded vectors corresponding to the object to be processed can be obtained, thereby enhancing the processing capability of the cold quantity determination model for the cold quantity characteristic data, improving the compatibility and adaptability of the cold quantity determination model to different types of characteristic data, and improving the accuracy and efficiency of the cold quantity determination model in predicting or controlling the cold quantity.

[0075] In some embodiments, each embedding vector corresponding to the object to be processed includes a category embedding vector corresponding to the object to be processed, a numerical embedding vector corresponding to the object to be processed, and a Boolean embedding vector corresponding to the object to be processed; the category feature data corresponding to the object to be processed, the numerical feature data corresponding to the object to be processed, and the Boolean feature data corresponding to the object to be processed are respectively vectorized and embedded to obtain each embedding vector corresponding to the object to be processed, including: splicing the category feature data to obtain field splicing data corresponding to the category feature data; embedding the field splicing data to obtain the category embedding vector corresponding to the object to be processed; encoding and embedding the numerical feature data to obtain the column name vector corresponding to the numerical feature data; normalizing the numerical feature data to obtain the normalized numerical value corresponding to the numerical feature data; multiplying the column name vector corresponding to the numerical feature data and the normalized numerical value corresponding to the numerical feature data to obtain the numerical embedding vector corresponding to the object to be processed; determining the Boolean embedding vector corresponding to the object to be processed according to the preset value conditions and the Boolean feature data corresponding to the object to be processed.

[0076] Specifically, multiple category fields in the category feature data of the object to be processed can be merged into a field spliced ​​data through splicing processing, and then the field spliced ​​data can be embedded to obtain a category embedding vector. The column name of the numerical feature data can be converted into a column name vector through encoding embedding processing. The column name vector can be used to represent different numerical attributes. The actual value in the numerical feature data can be normalized so that the obtained numerical value is in the same dimension. Then the column name vector can be multiplied by the normalized numerical value to obtain a numerical embedding vector. According to the preset value conditions, the Boolean feature data can be converted into a Boolean embedding vector. The Boolean embedding vector can be used to represent the specific state of the object to be processed.

[0077] The category embedding vector corresponding to the object to be processed may be a vector obtained through embedding processing, and the category embedding vector corresponding to the object to be processed may be used to capture association and difference information between different categories.

[0078] The numerical embedding vector corresponding to the object to be processed may be a vector obtained by multiplying the column name vector of the numerical feature data by the normalized numerical value.

[0079] The Boolean embedding vector corresponding to the object to be processed may be a vector obtained by converting Boolean feature data according to a preset value condition, and the Boolean embedding vector corresponding to the object to be processed may be used to characterize a specific state or condition of the object to be processed.

[0080] The field concatenated data corresponding to the category feature data may be a long character string or sequence obtained by concatenating multiple category feature fields of the object to be processed.

[0081] The column name vector corresponding to the numerical feature data may be a vector obtained by converting the column name of the numerical feature data through encoding embedding processing, and the column name vector corresponding to the numerical feature data may be used to identify different numerical attributes in the embedding vector.

[0082] The normalized value corresponding to the numerical feature data may be a result obtained by normalizing the actual value of the numerical feature data.

[0083] The preset value condition may be a rule or condition for converting Boolean feature data into a Boolean embedding vector, such as a mapping rule that a true value is mapped to 1 and a false value is mapped to 0.

[0084] For example, in a smart home system, the preferences of family members, the current room temperature, and the outdoor temperature can be obtained as cooling feature data, where the preferences of family members can be used as category feature data, and the category feature data can be processed by field concatenation and embedding to obtain a category embedding vector; the current room temperature and outdoor temperature can be used as numerical feature data, and the numerical feature data can be encoded, embedded and normalized to obtain a numerical embedding vector; whether family members are at home can be used as Boolean feature data, and according to preset value conditions, the value condition can be 1 for at home and 0 for not at home, and a Boolean embedding vector can be obtained by processing. The obtained embedding vector can be input into the trained cooling determination model, and the cooling determination model can predict the family members' demand for indoor temperature based on the embedded vector, and adjust the operating status of the indoor temperature control equipment accordingly.

[0085] According to the technical solution provided by the embodiment of the present disclosure, multiple category fields in the category feature data of the object to be processed are merged into one field spliced ​​data through splicing processing, and then the field spliced ​​data can be embedded to obtain a category embedding vector, and the column name of the numerical feature data can be converted into a column name vector through encoding embedding processing, and the actual value in the numerical feature data can be normalized so that the obtained numerical value is in the same dimension, and then the column name vector can be multiplied by the normalized numerical value to obtain a numerical embedding vector, and the Boolean feature data can be converted into a Boolean embedding vector according to preset value conditions, thereby enhancing the understanding and expression ability of the cooling determination model for complex feature data, improving the flexibility and accuracy of the cooling determination model in processing different types of feature data, and improving the performance of the cooling determination model in cooling prediction or control and other aspects.

[0086] In some embodiments, a gating mechanism is performed on each embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed, including: retaining each embedding vector corresponding to the object to be processed to obtain a retained embedding vector corresponding to the object to be processed; forgetting the retained embedding vector corresponding to the object to be processed to obtain a candidate hidden state vector corresponding to the object to be processed; and linearly interpolating the candidate hidden state vectors corresponding to the object to be processed and the retained embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed.

[0087] Specifically, each embedding vector corresponding to the processed object can be retained through the update gate of the gating mechanism to filter the degree to which the hidden state information of the previous moment is retained to the current moment, and the retained embedding vector is obtained. Then, the retained embedding vector can be forgotten through the reset gate in the gating mechanism to reduce or ignore unimportant information to obtain a candidate hidden state vector. The candidate hidden state vector and the retained embedding vector can be linearly interpolated and fused according to their weight or importance to obtain a hidden state vector.

[0088] Among them, the retained embedding vector corresponding to the object to be processed can be a vector obtained by processing through the update gate in the gating mechanism; the candidate hidden state vector corresponding to the object to be processed can be a vector obtained after forgetting processing through the reset gate in the gating mechanism.

[0089] For example, in an intelligent temperature control system, the embedded vector can be processed by a gating mechanism, can be processed by an update gate in the gating mechanism to obtain a retained embedded vector, can be processed by a reset gate in the gating mechanism to obtain a candidate hidden state vector, and then the candidate hidden state vector and the retained embedded vector can be linearly interpolated to obtain a hidden state vector.

[0090] According to the technical solution provided by the embodiment of the present disclosure, each embedded vector corresponding to the processed object is retained through the update gate of the gating mechanism to obtain a retained embedded vector, and then the retained embedded vector can be forgotten through the reset gate in the gating mechanism to obtain a candidate hidden state vector, and the candidate hidden state vector and the retained embedded vector can be linearly interpolated and fused according to their weights or importance to obtain a hidden state vector, thereby enhancing the screening and integration capabilities of the cold volume determination model for feature information, improving the adaptability and robustness of the cold volume determination model to complex data, and improving the performance of the cold volume determination model in cold volume prediction.

[0091] In some embodiments, the hidden state vector corresponding to the object to be processed is processed by a preset activation function to obtain the target cooling data corresponding to the object to be processed, including: activating the hidden state vector corresponding to the object to be processed to obtain the activation vector corresponding to the object to be processed; converting the activation vector corresponding to the object to be processed to obtain the target cooling data corresponding to the object to be processed.

[0092] Specifically, the hidden state vector corresponding to the object to be processed can be input into a preset activation function for nonlinear transformation processing to obtain the activation vector corresponding to the object to be processed. The preset activation function includes but is not limited to Sigmoid, ReLU, or Tanh, etc., which are not limited here. The activation vector can be transformed through a fully connected layer to adjust the dimension and numerical range of the activation vector to match the target cooling data to obtain the target cooling data corresponding to the object to be processed.

[0093] The activation vector corresponding to the object to be processed may be a vector obtained after being processed by a preset activation function, and the activation vector corresponding to the object to be processed may include information of the hidden state vector after nonlinear transformation.

[0094] For example, in an intelligent refrigeration system, the hidden state vector can be input into a Sigmoid function for activation processing to obtain an activation vector, which can then be transformed to obtain the target cooling capacity data corresponding to the cold storage room.

[0095] According to the technical solution provided by the embodiment of the present disclosure, the hidden state vector corresponding to the object to be processed is input into a preset activation function and nonlinear transformation processing is performed to obtain the activation vector corresponding to the object to be processed. The activation vector can be converted through a fully connected layer, and the dimension and numerical range of the activation vector are adjusted to match the target cooling data to obtain the target cooling data corresponding to the object to be processed, thereby enhancing the nonlinear processing capability of the cooling determination model, improving the fitting capability of the cooling determination model for complex data, and improving the accuracy and generalization capability of the cooling determination model in cooling prediction.

[0096] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present disclosure, and will not be described in detail here.

[0097] The following are embodiments of the device disclosed herein, which can be used to execute the method embodiments disclosed herein. For details not disclosed in the device embodiments disclosed herein, please refer to the method embodiments disclosed herein.

[0098] Figure 3 Schematic diagram of a cooling capacity determination device provided by an embodiment of the present disclosure. Figure 3 As shown, the device for determining the cooling capacity includes:

[0099] The first processing module 301 is used to input at least one cooling characteristic data corresponding to the object to be processed into the trained cooling determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed;

[0100] The second processing module 302 is used to perform gating mechanism processing on each embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed;

[0101] The third processing module 303 is used to process the hidden state vector corresponding to the object to be processed by a preset activation function to obtain the target cooling data corresponding to the object to be processed;

[0102] The fourth processing module 304 is used to send the target cooling capacity data corresponding to the object to be processed to the target terminal device for cooling capacity control.

[0103] According to the technical solution provided by the embodiment of the present disclosure, at least one cold feature data corresponding to the object to be processed is embedded through the embedding layer in the trained cold determination model to obtain each embedding vector corresponding to the object to be processed, and then the gated loop unit can be used to perform gating mechanism processing on each embedding vector corresponding to the object to be processed to obtain the hidden state vector corresponding to the object to be processed, and the hidden state vector corresponding to the object to be processed is activated and transformed based on a preset activation function, and the hidden state vector corresponding to the object to be processed is converted into target cold data corresponding to the object to be processed that is easy to read and understand, and then the target cold data corresponding to the object to be processed can be sent to the target terminal device for releasing and controlling the cold, thereby improving the accuracy of cold determination and enhancing the versatility of the cold determination model. The data is converted into a vector through the embedding layer, which improves the processing efficiency of the cold determination model, enhances the degree of understanding of the data, improves the scalability of the cold determination model, and ensures the real-time and effectiveness of cold determination.

[0104] In some embodiments, the above-mentioned cold capacity determination device is also used to obtain at least one historical cold capacity related data corresponding to the object to be processed; perform data type check processing on each historical cold capacity related data corresponding to the object to be processed, and obtain date data corresponding to the object to be processed and numerical feature data corresponding to the object to be processed; perform feature construction processing on the date data corresponding to the object to be processed, and obtain time data corresponding to the object to be processed; perform standardization processing on the numerical feature data corresponding to the object to be processed, and obtain standardized data corresponding to the object to be processed; format the time data corresponding to the object to be processed and the standardized data corresponding to the object to be processed, and obtain at least one cold capacity feature data corresponding to the object to be processed.

[0105] In some embodiments, the above-mentioned cold capacity determination device is also used to obtain cold capacity training data sets corresponding to multiple training objects, each cold capacity training data set contains multiple cold capacity training data and real cold capacity data corresponding to the training object; the cold capacity training data sets corresponding to each training object are respectively input into the cold capacity determination model for embedding processing to obtain the embedding vector set corresponding to each training object, and the embedding vector set corresponding to each training object contains multiple embedding vectors corresponding to the training object; the embedding vector set corresponding to each training object is respectively processed by the gating mechanism to obtain the hidden state vector corresponding to each training object; the hidden state vector corresponding to each training object is respectively processed by a preset activation function to obtain the training cold capacity data corresponding to each training object; the mean square error loss corresponding to the cold capacity determination model is determined according to the training cold capacity data corresponding to each training object and the real cold capacity data corresponding to each training object; the parameters in the cold capacity determination model are updated according to the mean square error loss corresponding to the cold capacity determination model through a cyclic iteration.

[0106] In some embodiments, the first processing module 301 is specifically used to perform data classification processing on at least one cold characteristic data corresponding to the object to be processed to obtain category characteristic data corresponding to the object to be processed, numerical characteristic data corresponding to the object to be processed, and Boolean characteristic data corresponding to the object to be processed; the category characteristic data corresponding to the object to be processed, the numerical characteristic data corresponding to the object to be processed, and the Boolean characteristic data corresponding to the object to be processed are respectively vectorized and embedded to obtain each embedded vector corresponding to the object to be processed.

[0107] In some embodiments, the category feature data corresponding to the object to be processed, the numerical feature data corresponding to the object to be processed, and the Boolean feature data corresponding to the object to be processed are respectively vectorized and embedded to obtain the embedding vectors corresponding to the object to be processed. The embedding vectors are specifically used to: splice the category feature data to obtain field splicing data corresponding to the category feature data; embed the field splicing data to obtain the category embedding vector corresponding to the object to be processed; encode and embed the numerical feature data to obtain the column name vector corresponding to the numerical feature data; normalize the numerical feature data to obtain the normalized numerical value corresponding to the numerical feature data; multiply the column name vector corresponding to the numerical feature data by the normalized numerical value corresponding to the numerical feature data to obtain the numerical embedding vector corresponding to the object to be processed; determine the Boolean embedding vector corresponding to the object to be processed according to the preset value conditions and the Boolean feature data corresponding to the object to be processed.

[0108] In some embodiments, the second processing module 302 is specifically used to retain each embedding vector corresponding to the object to be processed to obtain a retained embedding vector corresponding to the object to be processed; forget the retained embedding vector corresponding to the object to be processed to obtain a candidate hidden state vector corresponding to the object to be processed; perform linear interpolation processing on the candidate hidden state vector corresponding to the object to be processed and the retained embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed.

[0109] In some embodiments, the third processing module 303 is specifically used to activate the hidden state vector corresponding to the object to be processed to obtain the activation vector corresponding to the object to be processed; convert the activation vector corresponding to the object to be processed to obtain the target cooling data corresponding to the object to be processed.

[0110] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0111] Figure 4 Schematic diagram of an electronic device 4 provided in an embodiment of the present disclosure. Figure 4 As shown, the electronic device 4 of this embodiment includes: a processor 401, a memory 402, and a computer program 403 stored in the memory 402 and executable on the processor 401. When the processor 401 executes the computer program 403, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor 401 executes the computer program 403, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0112] The electronic device 4 may be a desktop computer, a notebook, a PDA, a cloud server, or other electronic device. The electronic device 4 may include, but is not limited to, a processor 401 and a memory 402. Those skilled in the art will appreciate that Figure 4 The electronic device 4 is merely an example and does not limit the electronic device 4 , and may include more or less components than those shown in the figure, or different components.

[0113] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0114] The memory 402 may be an internal storage unit of the electronic device 4, for example, a hard disk or memory of the electronic device 4. The memory 402 may also be an external storage device of the electronic device 4, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 4. The memory 402 may also include both an internal storage unit and an external storage device of the electronic device 4. The memory 402 is used to store computer programs and other programs and data required by the electronic device.

[0115] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units.

[0116] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium (e.g., a computer-readable storage medium). Based on this understanding, the present disclosure implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, and the computer program code may be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable storage medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0117] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should all be included in the protection scope of the present disclosure.

Claims

1. A method for determining cooling capacity, characterized in that: include: Inputting at least one cooling characteristic data corresponding to the object to be processed into the trained cooling determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed; Performing gating mechanism processing on each embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed; The hidden state vector corresponding to the object to be processed is processed by a preset activation function to obtain the target cooling data corresponding to the object to be processed; The target cooling capacity data corresponding to the object to be processed is sent to the target terminal device for cooling capacity control.

2. The method for determining cooling capacity according to claim 1, characterized in that: Before inputting at least one cooling characteristic data corresponding to the object to be processed into the trained cooling determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed, the method further includes: Acquire at least one historical cooling capacity related data corresponding to the object to be processed; Performing data type check processing on each historical cooling capacity related data corresponding to the object to be processed, and obtaining date data corresponding to the object to be processed and numerical feature data corresponding to the object to be processed; Performing feature construction processing on the date data corresponding to the object to be processed to obtain time data corresponding to the object to be processed; Performing standardization processing on the numerical feature data corresponding to the object to be processed to obtain standardized data corresponding to the object to be processed; The time data corresponding to the object to be processed and the standardized data corresponding to the object to be processed are formatted to obtain at least one cooling characteristic data corresponding to the object to be processed.

3. The method for determining cooling capacity according to claim 1, characterized in that: Before inputting at least one cooling characteristic data corresponding to the object to be processed into the trained cooling determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed, the method further includes: Acquire a plurality of cold capacity training data sets corresponding to the training objects, each of the cold capacity training data sets comprising a plurality of cold capacity training data and real cold capacity data corresponding to the training object; Inputting the cold volume training data sets corresponding to each of the training objects into the cold volume determination model for embedding processing, and obtaining the embedding vector sets corresponding to each of the training objects, wherein the embedding vector sets corresponding to each of the training objects include multiple embedding vectors corresponding to the training objects; Performing gating mechanism processing on the embedding vector sets corresponding to each of the training objects respectively to obtain the hidden state vectors corresponding to each of the training objects; The hidden state vectors corresponding to each of the training objects are processed respectively by the preset activation function to obtain the training cooling data corresponding to each of the training objects; Determine the mean square error loss corresponding to the cooling capacity determination model according to the training cooling capacity data corresponding to each of the training objects and the real cooling capacity data corresponding to each of the training objects; The parameters in the cooling capacity determination model are updated according to the mean square error loss corresponding to the cooling capacity determination model in a loop iteration manner.

4. The method for determining cooling capacity according to claim 1, characterized in that: The step of inputting at least one cooling characteristic data corresponding to the object to be processed into the trained cooling determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed includes: Performing data classification processing on at least one cooling characteristic data corresponding to the object to be processed to obtain category characteristic data corresponding to the object to be processed, numerical characteristic data corresponding to the object to be processed, and Boolean characteristic data corresponding to the object to be processed; Vectorized embedding processing is performed on the category feature data corresponding to the object to be processed, the numerical feature data corresponding to the object to be processed, and the Boolean feature data corresponding to the object to be processed, respectively, to obtain respective embedding vectors corresponding to the object to be processed.

5. The method for determining cooling capacity according to claim 4, characterized in that: Each embedding vector corresponding to the object to be processed includes a category embedding vector corresponding to the object to be processed, a numerical embedding vector corresponding to the object to be processed, and a Boolean embedding vector corresponding to the object to be processed; The vectorization embedding process is respectively performed on the category feature data corresponding to the object to be processed, the numerical feature data corresponding to the object to be processed, and the Boolean feature data corresponding to the object to be processed to obtain respective embedding vectors corresponding to the object to be processed, including: Performing splicing processing on the category characteristic data to obtain field splicing data corresponding to the category characteristic data; Embedding the field concatenated data to obtain a category embedding vector corresponding to the object to be processed; Performing encoding and embedding processing on the numerical feature data to obtain a column name vector corresponding to the numerical feature data; Normalizing the numerical feature data to obtain a normalized value corresponding to the numerical feature data; Multiplying the column name vector corresponding to the numerical feature data and the normalized numerical value corresponding to the numerical feature data to obtain a numerical embedding vector corresponding to the object to be processed; According to a preset value condition and the Boolean feature data corresponding to the object to be processed, a Boolean embedding vector corresponding to the object to be processed is determined.

6. The method for determining cooling capacity according to claim 1, characterized in that: The performing of a gating mechanism on each embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed includes: Performing retention processing on each embedding vector corresponding to the object to be processed to obtain a retained embedding vector corresponding to the object to be processed; Performing forgetting processing on the retained embedding vector corresponding to the object to be processed to obtain a candidate hidden state vector corresponding to the object to be processed; Linear interpolation processing is performed on the candidate hidden state vector corresponding to the object to be processed and the retained embedding vector corresponding to the object to be processed to obtain the hidden state vector corresponding to the object to be processed.

7. The method for determining cooling capacity according to claim 1, characterized in that: The step of processing the hidden state vector corresponding to the object to be processed by a preset activation function to obtain target cooling data corresponding to the object to be processed includes: Performing activation processing on the hidden state vector corresponding to the object to be processed to obtain an activation vector corresponding to the object to be processed; The activation vector corresponding to the object to be processed is converted to obtain the target cooling capacity data corresponding to the object to be processed.

8. A device for determining cooling capacity, characterized in that: include: A first processing module is used to input at least one cooling characteristic data corresponding to the object to be processed into the trained cooling determination model for embedding processing to obtain each embedding vector corresponding to the object to be processed; A second processing module is used to perform gating mechanism processing on each embedding vector corresponding to the object to be processed to obtain a hidden state vector corresponding to the object to be processed; A third processing module is used to process the hidden state vector corresponding to the object to be processed by a preset activation function to obtain the target cooling data corresponding to the object to be processed; The fourth processing module is used to send the target cooling capacity data corresponding to the object to be processed to the target terminal device for cooling capacity control.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images