Intelligent liquid nitrogen storage tank

Abstract

Disclosed is an intelligent liquid nitrogen storage tank. The intelligent liquid nitrogen storage tank comprises: a container for storing liquid nitrogen; a liquid nitrogen inlet for delivering liquid nitrogen into the container; a liquid nitrogen outlet for discharging liquid nitrogen from the container; a pressure sensor disposed within the container for detecting the pressure within the container; a temperature sensor disposed within the container for detecting the temperature within the container; a valve; and a controller communicatively connected to the pressure sensor and temperature sensor for controlling the valve opening. This allows for automated management of liquid nitrogen storage, reducing the risk of manual intervention and misoperation.

Description

Technical Field

[0001] The present application relates to the field of liquid nitrogen storage, and more specifically, to an intelligent liquid nitrogen storage tank. Background Art

[0002] Liquid nitrogen is a cryogenic liquid widely used in scientific research, industrial production, and medical fields. During its storage and use, temperature and pressure stability are crucial to ensuring its quality and safety. Liquid nitrogen's temperature stability directly affects its physical and chemical properties, while pressure stability is crucial to the safety of its containers.

[0003] Traditional liquid nitrogen storage tanks typically maintain a low temperature by reducing heat conduction from the outside through insulation and reducing the effects of thermal radiation through a vacuum layer. However, when faced with changes in the external environment and internal pressure fluctuations, traditional storage tanks have difficulty maintaining stable temperature and pressure of the liquid nitrogen. This can lead to accelerated evaporation of the liquid nitrogen, affecting its efficiency and quality.

[0004] Furthermore, traditional liquid nitrogen storage tanks typically require manual monitoring and adjustment to maintain stable liquid nitrogen temperature and pressure. This reliance on manual operation can easily lead to human error or inadequate monitoring, increasing safety risks and uncertainty in liquid nitrogen quality. Furthermore, traditional liquid nitrogen storage tanks lack intelligent management and automatic adjustment capabilities. They are unable to monitor liquid nitrogen temperature and pressure in real time and automatically adjust them as needed, resulting in inefficient management and wasted resources.

[0005] Furthermore, therefore, a smart liquid nitrogen storage tank is desired. Summary of the Invention

[0006] To address the above technical issues, the present application is proposed. Embodiments of the present application provide an intelligent liquid nitrogen storage tank that can automate the management of liquid nitrogen storage, reducing the risk of manual intervention and misoperation.

[0007] According to one aspect of the present application, there is provided an intelligent liquid nitrogen storage tank, comprising:

[0008] Container for storing liquid nitrogen;

[0009] a liquid nitrogen inlet, for delivering liquid nitrogen into the container;

[0010] a liquid nitrogen outlet, for discharging liquid nitrogen from the container;

[0011] A pressure sensor disposed inside the container, for detecting the pressure inside the container;

[0012] A temperature sensor disposed inside the container, for detecting the temperature inside the container;

[0013] Valves; and

[0014] A controller is communicatively connected to the pressure sensor and the temperature sensor, and is used to control the opening value of the valve.

[0015] Compared to the prior art, the intelligent liquid nitrogen storage tank provided in this application includes: a container for storing liquid nitrogen; a liquid nitrogen inlet for delivering liquid nitrogen into the container; a liquid nitrogen outlet for discharging liquid nitrogen from the container; a pressure sensor disposed within the container for detecting the pressure within the container; a temperature sensor disposed within the container for detecting the temperature within the container; a valve; and a controller communicatively connected to the pressure sensor and temperature sensor for controlling the valve opening. This allows for automated management of liquid nitrogen storage, reducing the risk of manual intervention and misoperation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for the description of the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present application.

[0017] Figure 1 Schematic diagram of an intelligent liquid nitrogen storage tank according to an embodiment of the present application.

[0018] Figure 2 Schematic diagram of a block diagram of the controller in the intelligent liquid nitrogen storage tank according to an embodiment of the present application.

[0019] Figure 3 Schematic diagram of a block diagram of the intra-container parameter timing collaborative analysis module in the intelligent liquid nitrogen storage tank according to an embodiment of the present application.

[0020] Figure 4 Schematic diagram of the valve opening control module in the intelligent liquid nitrogen storage tank according to an embodiment of the present application.

[0021] Figure 5 The figure is a flow chart of a method for an intelligent liquid nitrogen storage tank according to an embodiment of the present application.

[0022] Figure 6 Schematic diagram of the system architecture of the method for an intelligent liquid nitrogen storage tank according to an embodiment of the present application.

[0023] Figure 7 This is an application scenario diagram of the intelligent liquid nitrogen storage tank according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without making creative efforts also fall within the scope of protection of this application.

[0025] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0026] Although the present application makes various references to certain modules in the system according to 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 illustrative only, and different aspects of the system and method can use different modules.

[0027] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[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 herein.

[0029] To address the above technical issues, the technical concept of this application is to monitor and collect the pressure and temperature values within the container in real time through pressure sensors and temperature sensors deployed inside the container, and introduce data processing and analysis algorithms on the back end to perform time-series collaborative analysis of the pressure and temperature values within the container, thereby automatically adjusting the valve opening value in real time. In this way, it is possible to achieve automated management of liquid nitrogen storage, reduce the risk of manual intervention and misoperation. At the same time, it is possible to maintain a constant temperature and pressure of liquid nitrogen, improve the efficiency and quality of liquid nitrogen use, and avoid safety hazards or waste.

[0030] Figure 1 FIG. 1 is a block diagram of an intelligent liquid nitrogen storage tank according to an embodiment of the present application. Figure 1As shown, the intelligent liquid nitrogen storage tank 10 according to an embodiment of the present application includes: a container 11 for storing liquid nitrogen; a liquid nitrogen inlet 12 for transporting liquid nitrogen into the container; a liquid nitrogen outlet 13 for discharging liquid nitrogen from the container; a pressure sensor 14 provided inside the container for detecting the pressure inside the container; a temperature sensor 15 provided inside the container for detecting the temperature inside the container; a valve 16; and a controller 100, which can be communicatively connected to the pressure sensor 14 and the temperature sensor 15 to control the opening value of the valve 16.

[0031] It should be understood that in the intelligent liquid nitrogen storage tank 10, the container 11 is used to store the main part of the liquid nitrogen; the liquid nitrogen inlet 12 is used to transport liquid nitrogen into the container and inject the liquid nitrogen into the storage tank; the liquid nitrogen outlet 13 is used to discharge liquid nitrogen from the container, and the liquid nitrogen can be taken out or discharged through the outlet; the pressure sensor 14 is arranged inside the container to detect the pressure change in the container. The pressure sensor can monitor the pressure level inside the liquid nitrogen storage tank in real time; the temperature sensor 15 is arranged inside the container to detect the temperature change in the container. The temperature sensor can monitor the temperature level inside the liquid nitrogen storage tank in real time; the valve 16 is used to control the flow rate of liquid nitrogen in and out of the container, and the opening value of the valve can adjust the flow rate of liquid nitrogen; the controller 100: The controller is communicated with the pressure sensor and the temperature sensor to monitor and control the state of the liquid nitrogen storage tank. The controller can control the flow rate of liquid nitrogen in and out of the container by adjusting the opening value of the valve according to the feedback information of the pressure and temperature sensors, thereby realizing intelligent control of the liquid nitrogen storage tank. By monitoring and regulating the pressure and temperature inside the liquid nitrogen storage tank, the controller ensures safe and stable operation. Based on sensor feedback, the controller automatically adjusts the valve opening to achieve precise control of liquid nitrogen in and out of the container, thereby meeting the storage tank's operating requirements.

[0032] Furthermore, Figure 2 FIG. 1 is a block diagram of the controller in the intelligent liquid nitrogen storage tank according to an embodiment of the present application. Figure 2As shown, according to the intelligent liquid nitrogen storage tank of an embodiment of the present application, the controller 100 includes: a data acquisition module 110, which is used to obtain the pressure value and the temperature value in the container at multiple predetermined time points within a predetermined time period collected by the pressure sensor and the temperature sensor; a data parameter time series arrangement module 120, which is used to arrange the pressure value and the temperature value in the container at the multiple predetermined time points into a container pressure time series input vector and a container temperature time series input vector according to the time dimension; a container parameter time series collaborative analysis module 130, which is used to perform time series correlation encoding on the container pressure time series input vector and the container temperature time series input vector to obtain a container temperature-pressure time series correlation feature; and a valve opening control module 140, which is used to determine whether the valve opening value at the current time point should be increased or decreased based on the container temperature-pressure time series correlation feature.

[0033] Specifically, in the technical solution of the present application, first, the pressure values and temperature values within the container at multiple predetermined time points within a predetermined time period, collected by a pressure sensor and a temperature sensor, are acquired. Next, considering that the pressure values and temperature values within the container both have a time-series dynamic change pattern in the time dimension, in order to establish a time series model of these parameters so as to analyze the state of the liquid nitrogen storage tank, in the technical solution of the present application, the pressure values and temperature values within the container at the multiple predetermined time points are arranged according to the time dimension into a container pressure time series input vector and a container temperature time series input vector, thereby integrating the distribution information of the pressure values and temperature values within the container in the time series, respectively.

[0034] Then, considering that in a liquid nitrogen storage tank, the pressure and temperature inside the container are often correlated, their temporal changes will affect each other. In other words, the pressure value inside the container and the temperature value inside the container not only have temporal variation characteristics in the time dimension, but also have a temporal synergistic correlation between the two, jointly affecting the storage state of the liquid nitrogen. Therefore, in the technical solution of the present application, it is necessary to perform a temporal synergistic analysis of the pressure value inside the container and the temperature value inside the container, so as to adaptively control the valve opening value to maintain a constant temperature and pressure of the liquid nitrogen.

[0035] Based on this, the technical solution of the present application further performs association encoding on the container pressure time series input vector and the container temperature time series input vector to obtain a container pressure-temperature time series association matrix, thereby reflecting the correlation between the pressure and temperature within the container in the time dimension. Subsequently, the container pressure-temperature time series association matrix is subjected to feature mining in a container temperature-pressure time series association feature extractor based on a convolutional neural network model to extract time series collaborative association feature information between the container pressure values and the container temperature values in the time dimension, thereby obtaining a container temperature-pressure time series association feature vector.

[0036] In practical applications, due to noise in sensor data, equipment failures, or other interfering factors, the time series data of the temperature and pressure inside a container may contain some unnecessary noise or outliers. This noise can negatively impact subsequent analysis and prediction, necessitating denoising. Based on this, in the technical solution of the present application, the time series correlation feature vector of the temperature and pressure inside the container is passed through a feature denoiser based on a Bi-LSTM model to obtain a denoised time series correlation feature vector of the temperature and pressure inside the container. It should be understood that a bidirectional long short-term memory network is a deep learning model suitable for sequence data processing that can effectively capture long-term dependencies and contextual information in time series data. Therefore, by using the Bi-LSTM model as a feature denoiser, its powerful modeling capabilities can be leveraged to learn and extract valid information from the time series correlation feature vector of the temperature and pressure inside the container, while suppressing the impact of noise and outliers. Specifically, the Bi-LSTM model can simultaneously consider the time series correlation contextual information about the temperature and pressure inside the container before and after the current moment, modeling the feature dependencies of the temperature and pressure time series data through forward and backward memory units. During the feature denoising process, the Bi-LSTM model can learn the potential patterns and regularities in the temperature-pressure time series correlation feature vector in the container and eliminate noise and outliers.

[0037] Accordingly, if Figure 3As shown, the container parameter time series collaborative analysis module 130 includes: a container temperature-pressure time series correlation encoding unit 131, configured to perform correlation encoding on the container pressure time series input vector and the container temperature time series input vector to obtain a container pressure-temperature time series correlation matrix; a temperature-pressure time series correlation feature extraction unit 132, configured to perform feature extraction on the container pressure-temperature time series correlation matrix using a container temperature-pressure time series correlation feature extractor based on a deep neural network model to obtain a container temperature-pressure time series correlation feature vector; and a feature denoising unit 133, configured to perform feature denoising on the container temperature-pressure time series correlation feature vector to obtain a denoised container temperature-pressure time series correlation feature vector as the container temperature-pressure time series correlation feature. It should be understood that the container parameter time series collaborative analysis module 130 includes three units: the container temperature-pressure time series correlation encoding unit 131, the temperature-pressure time series correlation feature extraction unit 132, and the feature denoising unit 133. The function of the temperature-pressure time series correlation coding unit 131 within the container is to correlate the time series data of temperature and pressure for subsequent feature extraction and analysis. The function of the temperature-pressure time series correlation feature extraction unit 132 is to capture the patterns and regularities of temperature and pressure changes by learning and extracting the correlation features between the temperature and pressure within the container. The function of the feature denoising unit 133 is to remove noise or redundant information from the feature vector to improve the quality and reliability of the features. These units work together in the collaborative analysis module of the time series of parameters within the container. Through steps such as correlation coding, feature extraction, and feature denoising, correlation features are extracted from the time series data of temperature and pressure within the container for subsequent analysis and application.

[0038] More specifically, in the temperature-pressure time series correlation feature extraction unit 132, the deep neural network model is a convolutional neural network model. It is worth mentioning that a convolutional neural network (CNN) is a deep learning model specifically designed for processing data with a grid structure. A CNN is composed of convolutional layers, pooling layers, and fully connected layers. The convolutional layer is the core component of a CNN. It performs a convolution operation on the input data by applying a series of learnable filters (also known as convolution kernels) to extract local features from the input data. The convolution operation can capture local spatial relationships in an image, such as edges, textures, and other features. The pooling layer is used to reduce the spatial dimension of the feature map, reducing the number of parameters while retaining important features. Common pooling operations include max pooling and average pooling. Finally, the fully connected layer connects the outputs of the convolutional and pooling layers to one or more fully connected layers for tasks such as classification or regression. Convolutional neural networks are trained using a backpropagation algorithm to optimize network parameters, enabling them to automatically extract features from input data and perform well in various tasks. In the temperature-pressure time series correlation feature extraction unit 132, a feature extractor based on a convolutional neural network model is used to effectively learn and extract the correlation features between the temperature and pressure in the container.

[0039] More specifically, the feature denoising unit 133 is configured to pass the container temperature-pressure time series correlation feature vector through a feature denoiser based on a Bi-LSTM model to obtain the denoised container temperature-pressure time series correlation feature vector. It's worth noting that Bi-LSTM stands for Bidirectional Long Short-Term Memory Network, a variant of the Recurrent Neural Network (RNN) used to process sequential data, such as time series data or natural language text. LSTM is a special RNN unit that addresses the vanishing and exploding gradient problems in traditional RNNs by introducing a gating mechanism. An LSTM unit has a memory cell and three gates: an input gate, a forget gate, and an output gate. These gates control the flow of information, enabling the LSTM to selectively remember or forget past information, thereby better handling long-term dependencies. Bidirectional LSTM (Bi-LSTM) involves running two independent LSTM networks in each direction of the time series: one forward-to-backward and the other backward-to-forward. In this way, Bi-LSTM can simultaneously capture past and future contextual information, enabling a more comprehensive understanding of patterns and regularities in sequential data. In feature denoising unit 133, a feature denoiser based on the Bi-LSTM model is used to process the feature vectors of the container's internal temperature-pressure time series correlation. The Bi-LSTM model can learn long-term dependencies in time series data and denoise the feature vectors, removing noise or redundant information. Applying the Bi-LSTM model improves the quality and reliability of the feature vectors, resulting in a denoised feature vector of the container's internal temperature-pressure time series correlation for subsequent analysis and application.

[0040] Furthermore, the denoised container temperature-pressure time-series correlation feature vector is passed through a classifier to obtain a classification result. The classification result indicates whether the valve opening value at the current time point should be increased or decreased. In other words, the classification process is performed using the global time-series synergistic correlation features of the container temperature and pressure, thereby automatically adjusting the valve opening value in real time. This approach enables automated management of liquid nitrogen storage, reducing the risk of manual intervention and misoperation. At the same time, it maintains a constant temperature and pressure of liquid nitrogen, improving its efficiency and quality while avoiding potential safety hazards and waste.

[0041] Accordingly, if Figure 4As shown, the valve opening control module 140 includes: a feature gain optimization unit 141 for performing distribution gain on the de-noised container temperature-pressure time series correlation feature vector based on a probability density feature modeling paradigm to obtain a gain-de-noised container temperature-pressure time series correlation feature vector; and a liquid nitrogen parameter control unit 142 for passing the gain-de-noised container temperature-pressure time series correlation feature vector through a classifier to obtain a classification result. The classification result indicates whether the valve opening value at the current time point should be increased or decreased. It should be understood that the valve opening control module 140 includes the feature gain optimization unit 141 and the liquid nitrogen parameter control unit 142. The feature gain optimization unit 141 processes the de-noised container temperature-pressure time series correlation feature vector to obtain a gain-de-noised container temperature-pressure time series correlation feature vector based on a distribution gain based on a probability density feature modeling paradigm. This process can be achieved by adjusting the weights of each feature in the feature vector. By optimizing the feature gain, the expressive power of the feature vector can be further improved, allowing it to more accurately reflect the relationship between temperature and pressure in the container. The liquid nitrogen parameter control unit 142 uses a classifier to classify the time-series correlation feature vector of the temperature and pressure in the container after gain denoising to obtain a classification result. This classification result is used to indicate whether the valve opening value at the current time point should be increased or decreased. Based on the classification result of the feature vector, the valve opening can be controlled. By monitoring and analyzing the time-series correlation characteristics of the temperature and pressure in the container, the liquid nitrogen parameter control unit can intelligently adjust the valve opening according to the current state and target requirements to achieve precise control of the liquid nitrogen parameters. In summary, the feature gain optimization unit and the liquid nitrogen parameter control unit work together in the valve opening control module to achieve intelligent control of the valve opening by optimizing the feature vector and classification results to meet the requirements of the liquid nitrogen parameters in the container.

[0042] In particular, in the technical solution of the present application, in the process of passing the temperature-pressure time series correlation feature vector in the container through a feature denoiser based on a Bi-LSTM model to obtain a denoised temperature-pressure time series correlation feature vector in the container, the feature denoiser based on the Bi-LSTM model performs short-range-long-range bidirectional contextual correlation encoding based on the convolution kernel scale in the global time domain on the temperature-pressure time series correlation feature vector in the container, so as to perform feature denoising through the feature logical association between the local sequence feature distributions in the temperature-pressure time series correlation feature vector in the container. However, compared with the time series correlation features in the local time domain as foreground object features, when performing the contextual association representation between local time domains in the global time domain, background distribution noise related to the interference of the time series correlation feature distribution in each local time domain will also be introduced. Therefore, it is expected to enhance its expression effect based on the distribution characteristics of the denoised temperature-pressure time series correlation feature vector in the container.

[0043] Therefore, the applicant of the present application performs distribution gain based on the probability density feature imitation paradigm on the denoised temperature-pressure time series correlation feature vector in the container.

[0044] Accordingly, in a specific example, the feature gain optimization unit 141 is configured to perform a distribution gain based on a probability density feature imitation paradigm on the de-noised container temperature-pressure time series correlation feature vector using the following optimization formula to obtain the gain de-noised container temperature-pressure time series correlation feature vector; wherein the optimization formula is:

[0045]

[0046] Where V is the time series correlation feature vector of temperature and pressure in the container after denoising, v i is the eigenvalue of the ith position of the denoised temperature-pressure time series correlation feature vector, L is the length of the denoised temperature-pressure time series correlation feature vector, represents the square of the second norm of the denoised temperature-pressure time series correlation eigenvector in the container, α is a weighted hyperparameter, exp(·) represents the exponential operation of a value, and the exponential operation of the value represents the calculation of the natural exponential function value with the value as the power. is the eigenvalue of the i-th position of the temperature-pressure time series correlation eigenvector in the container after the gain denoising.

[0047] Here, based on the feature imitation paradigm of the standard Cauchy distribution for the natural Gaussian distribution on the probability density, the distribution gain based on the probability density feature imitation paradigm can use the feature scale as an imitation mask to distinguish the foreground object features and the background distribution noise in the high-dimensional feature space, thereby performing a distribution soft matching of the semantic cognition of the feature space mapping of the high-dimensional space based on the spatial hierarchical semantics of the high-dimensional features to obtain the unconstrained distribution gain of the high-dimensional feature distribution, thereby improving the expression effect of the temperature-pressure time series correlation feature vector in the container after denoising based on the feature distribution characteristics, thereby improving the accuracy of the classification result obtained by the classifier of the temperature-pressure time series correlation feature vector in the container after denoising, thereby improving the adjustment accuracy and adaptability of the valve opening value. In this way, the adaptive control of the valve opening value can be achieved based on the coordinated changes of the temperature and pressure in the actual container to realize intelligent liquid nitrogen storage and management, thereby ensuring the constant temperature and pressure of liquid nitrogen storage, improving the efficiency and quality of liquid nitrogen use, and avoiding safety hazards or waste, ensuring the stable storage and use of liquid nitrogen.

[0048] Furthermore, the liquid nitrogen parameter control unit 142 is used to: use the fully connected layer of the classifier to fully connect the temperature-pressure time series correlation feature vector in the container after gain denoising to obtain a coded classification feature vector; and input the coded classification feature vector into the Softmax classification function of the classifier to obtain the classification result.

[0049] It should be understood that the role of a classifier is to use given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression and SVM are commonly used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-class classification. However, this is prone to errors and is inefficient. A commonly used multi-classification method is the Softmax classification function.

[0050] That is, in the technical solution disclosed in the present invention, the labels of the classifier include the valve opening value at the current time point should increase (first label), and the valve opening value at the current time point should decrease (second label), wherein the classifier determines to which classification label the temperature-pressure time series correlation feature vector in the container after gain denoising belongs through the soft maximum function. It is worth noting that the first label p1 and the second label p2 here do not contain artificially set concepts. In fact, during the training process, the computer model does not have the concept of "the valve opening value at the current time point should increase or decrease". It only has two classification labels and the probability of the output feature under these two classification labels, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the valve opening value at the current time point should increase or decrease is actually converted from the classification label to a binary classification class probability distribution that conforms to natural laws. In essence, what is used is the physical meaning of the natural probability distribution of the label, rather than the linguistic text meaning of "the valve opening value at the current time point should increase or decrease".

[0051] It's worth noting that fully connected encoding involves linearly transforming and nonlinearly activating the input data through a fully connected layer to produce an encoded feature vector. In the valve opening control module, the classifier of the liquid nitrogen parameter control unit uses a fully connected layer to fully encode the gain-denoised, in-vessel temperature-pressure time series correlation feature vector. A fully connected layer consists of multiple neurons, each connected to all neurons in the previous layer. By learning weights and biases, the fully connected layer performs linear transformation and nonlinear mapping on the input feature vector. Fully connected encoding extracts high-level representations and abstract features from the feature vector to better convey the key information of the input data. Through fully connected encoding, the original gain-denoised, in-vessel temperature-pressure time series correlation feature vector is converted into a more expressive encoded classification feature vector. These encoded feature vectors can be better used for classification tasks, improving classifier performance and accuracy. In the classifier, the encoded classification feature vector is input into a softmax classification function, which normalizes the encoded classification feature vector and converts it into a classification result representing the probability distribution of different classes. The Softmax function determines the probability that the valve opening value at the current time point should increase or decrease, thereby achieving automatic control of the valve opening. In other words, the fully connected encoding extracts and encodes a high-level representation of the input feature vector through linear transformations and nonlinear activations in the fully connected layer to better support classification tasks.

[0052] In summary, the intelligent liquid nitrogen storage tank based on the embodiment of the present application is explained, which can realize the automated management of liquid nitrogen storage and reduce the risks of manual intervention and misoperation.

[0053] As described above, the smart liquid nitrogen storage tank according to the embodiment of the present application can be implemented in various terminal devices, such as a server having an algorithm based on the smart liquid nitrogen storage tank according to the embodiment of the present application. In one example, the smart liquid nitrogen storage tank according to the embodiment of the present application can be integrated into the terminal device as a software module and / or hardware module. For example, the smart liquid nitrogen storage tank according to the embodiment of the present application can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the smart liquid nitrogen storage tank according to the embodiment of the present application can also be one of the many hardware modules of the terminal device.

[0054] Alternatively, in another example, the smart liquid nitrogen storage tank based on the embodiment of the present application and the terminal device may also be separate devices, and the smart liquid nitrogen storage tank may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0055] Figure 5 The figure is a flow chart of a method for an intelligent liquid nitrogen storage tank according to an embodiment of the present application. Figure 6 Schematic diagram of the system architecture of the method for intelligent liquid nitrogen storage tank according to the embodiment of the present application. Figure 5 and Figure 6 As shown, according to the method of the intelligent liquid nitrogen storage tank of the embodiment of the present application, it includes: S110, obtaining the pressure value and the temperature value in the container at multiple predetermined time points within a predetermined time period collected by the pressure sensor and the temperature sensor; S120, arranging the pressure value and the temperature value in the container at the multiple predetermined time points into a container pressure time series input vector and a container temperature time series input vector according to the time dimension; S130, performing time series correlation encoding on the container pressure time series input vector and the container temperature time series input vector to obtain a container temperature-pressure time series correlation feature; and, S140, determining whether the valve opening value at the current time point should be increased or decreased based on the container temperature-pressure time series correlation feature.

[0056] In a specific example, in the above-mentioned method of the intelligent liquid nitrogen storage tank, the pressure time series input vector in the container and the temperature time series input vector in the container are time-series associated with each other to obtain the temperature-pressure time series associated features in the container, including: associating the pressure time series input vector in the container and the temperature time series input vector in the container to obtain the pressure-temperature time series associated matrix in the container; performing feature extraction on the pressure-temperature time series associated matrix in the container through a temperature-pressure time series associated feature extractor in the container based on a deep neural network model to obtain the temperature-pressure time series associated feature vector in the container; and, performing feature denoising on the temperature-pressure time series associated feature vector in the container to obtain the denoised temperature-pressure time series associated feature vector in the container as the temperature-pressure time series associated feature in the container.

[0057] Here, those skilled in the art will appreciate that the specific operations of each step in the method for the intelligent liquid nitrogen storage tank have been described above with reference to Figures 2 to 4 The description of the intelligent liquid nitrogen storage tank has been introduced in detail, and therefore, its repeated description will be omitted.

[0058] Figure 7 This is an application scenario diagram of the intelligent liquid nitrogen storage tank according to the embodiment of the present application. Figure 7 As shown, in this application scenario, first, the pressure values in the container at multiple predetermined time points within a predetermined time period collected by the pressure sensor 14 and the temperature sensor 15 are obtained (for example, Figure 7 D1) and the temperature value in the container (e.g. Figure 7Then, the pressure values and temperature values in the container at the plurality of predetermined time points are input into a server in which an algorithm of a smart liquid nitrogen storage tank is deployed (for example, Figure 7 S) as shown in , wherein the server is capable of using the algorithm of the intelligent liquid nitrogen storage tank to process the pressure values and temperature values in the container at the multiple predetermined time points to obtain a classification result indicating whether the valve opening value at the current time point should be increased or decreased.

[0059] This application uses specific terms to describe the embodiments of this application. For example, "first / second embodiment", "one embodiment", and / or "some embodiments" refer to a certain feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or multiple times in different places in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application may be appropriately combined.

[0060] In addition, it will be understood by those skilled in the art that various aspects of the present application can be illustrated and described by a number of patentable categories or situations, including any new and useful process, machine, product or combination of substances, or any new and useful improvements thereto. Accordingly, various aspects of the present application can be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may all be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". In addition, various aspects of the present application may be represented as a computer product located in one or more computer-readable media, which includes computer-readable program code.

[0061] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology and should not be interpreted in an idealized or highly formal sense, unless expressly defined as such herein.

[0062] The above is an illustration of the present invention and should not be considered as limiting thereof. Although several exemplary embodiments of the present invention have been described, it will be readily understood by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present invention. Therefore, all such modifications are intended to be included within the scope of the present invention as defined by the claims. It should be understood that the above is an illustration of the present invention and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present invention is defined by the claims and their equivalents.

Claims

1. An intelligent liquid nitrogen storage tank, characterized in that: include: Container for storing liquid nitrogen; a liquid nitrogen inlet, for delivering liquid nitrogen into the container; a liquid nitrogen outlet, for discharging liquid nitrogen from the container; A pressure sensor disposed inside the container, for detecting the pressure inside the container; A temperature sensor disposed inside the container, for detecting the temperature inside the container; valve; as well as a controller, the controller being communicatively connected to the pressure sensor and the temperature sensor, and configured to control the opening value of the valve; The controller includes: a data acquisition module, configured to acquire pressure values and temperature values within the container at a plurality of predetermined time points within a predetermined time period acquired by the pressure sensor and the temperature sensor; a data parameter time series arrangement module, configured to arrange the container pressure values and the container temperature values at the plurality of predetermined time points into a container pressure time series input vector and a container temperature time series input vector according to a time dimension; a container parameter time series collaborative analysis module, configured to perform time series correlation coding on the container pressure time series input vector and the container temperature time series input vector to obtain a container temperature-pressure time series correlation feature, wherein a container temperature-pressure time series correlation feature vector is obtained based on the container pressure time series input vector and the container temperature time series input vector, and feature denoising is performed on the container temperature-pressure time series correlation feature vector to obtain a denoised container temperature-pressure time series correlation feature vector as the container temperature-pressure time series correlation feature; and A valve opening control module, configured to determine whether the valve opening value at a current time point should be increased or decreased based on a time series correlation characteristic of temperature and pressure in the container; The valve opening control module includes: a feature gain optimization unit, configured to perform a distribution gain based on a probability density feature imitation paradigm on the de-noised temperature-pressure time series correlation feature vector in the container to obtain a gain-denoised temperature-pressure time series correlation feature vector in the container; a liquid nitrogen parameter control unit, configured to pass the gain-denoised temperature-pressure time series correlation feature vector in the container through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the valve opening value at the current time point should be increased or decreased; The characteristic gain optimization unit is used to: Performing a distribution gain based on a probability density feature imitation paradigm on the denoised temperature-pressure time series correlation feature vector in the container using the following optimization formula to obtain the gain denoised temperature-pressure time series correlation feature vector in the container; Wherein, the optimization formula is: in, is the time series correlation feature vector of temperature and pressure in the container after denoising, is the first eigenvector of the temperature-pressure time series correlation feature vector in the container after denoising. The eigenvalues at the positions, is the length of the temperature-pressure time series correlation feature vector in the container after denoising, represents the square of the second norm of the temperature-pressure time series correlation eigenvector in the container after denoising, and is a weighted hyperparameter, represents an exponential operation of a value, wherein the exponential operation of the value represents calculating the value of a natural exponential function with the value as a power, is the first eigenvector of the temperature-pressure time series correlation feature vector in the container after gain denoising. The eigenvalues at each position.

2. The intelligent liquid nitrogen storage tank according to claim 1, characterized in that: The container parameter timing collaborative analysis module includes: a container internal temperature-pressure time series correlation coding unit, configured to perform correlation coding on the container internal pressure time series input vector and the container internal temperature time series input vector to obtain a container internal pressure-temperature time series correlation matrix; a temperature-pressure time series correlation feature extraction unit, configured to extract features from the pressure-temperature time series correlation matrix within the container using a temperature-pressure time series correlation feature extractor within the container based on a deep neural network model to obtain a temperature-pressure time series correlation feature vector within the container; and The feature denoising unit is used to perform feature denoising on the temperature-pressure time series correlation feature vector in the container to obtain a denoised temperature-pressure time series correlation feature vector in the container as the temperature-pressure time series correlation feature in the container.

3. The intelligent liquid nitrogen storage tank according to claim 2, characterized in that: The deep neural network model is a convolutional neural network model.

4. The intelligent liquid nitrogen storage tank according to claim 3, characterized in that: The feature denoising unit is used to: The temperature-pressure time series correlation feature vector in the container is passed through a feature denoiser based on a Bi-LSTM model to obtain the denoised temperature-pressure time series correlation feature vector in the container.

5. The intelligent liquid nitrogen storage tank according to claim 4, characterized in that: The liquid nitrogen parameter control unit is used to: Using the fully connected layer of the classifier to perform fully connected encoding on the temperature-pressure time series correlation feature vector in the container after gain denoising to obtain an encoded classification feature vector; as well as The encoded classification feature vector is input into the Softmax classification function of the classifier to obtain the classification result.

Images