Ammonia adsorption tower driving control device and control method using the same

Abstract

To provide an artificial intelligence-based ammonia adsorption tower driving control device and a control method using the same, capable of contributing to environmental friendliness and producing clean energy by executing an abnormal situation processing process when an abnormal state occurs.SOLUTION: A control device 100 includes: a sensor unit 110 configured to measure an internal state of an adsorption tower; a memory configured to store one or more instructions; a processor configured to execute the one or more instructions stored in the memory; and a driving unit configured to drive the adsorption tower according to an adsorption cycle and a desorption cycle set on the basis of the internal state of the adsorption tower. Therein the processor is configured to: output an adsorption cycle and desorption cycle according to sensing data of the sensor unit using a trained artificial intelligence model; when the sensing data is within a preset optimal range, transmit a command to drive the adsorption tower according to the output adsorption cycle and desorption cycle to the driving unit; and if the sensing data exceeds the optimal range, transmit a command to execute an abnormal situation processing process according to a preset exception adsorption cycle and exception desorption cycle to the driving unit.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present disclosure relates to an ammonia absorber drive control device and a control method using the same. [Background technology]

[0002] Ammonia is one of the substances that causes bad odors and can be produced not only in livestock barns, but also in composting facilities that use chemical fertilizers and synthetic fibers, factories, sewage treatment plants, and other places.

[0003] Conventionally, an absorption tower is operated to adsorb and remove ammonia from air input into the absorption tower, and when the ammonia is adsorbed to a certain level or more and the removal efficiency of the absorption tower decreases, the ammonia is desorbed and the absorption tower is controlled so as to maintain a state in which it can be adsorbed again.

[0004] In order to repeat the above process, it is very important to control the adsorption / desorption cycle so that the adsorption and desorption processes are performed at the appropriate time. Conventionally, the vapor recovery method was used to desorb the VOCs adsorbed on the adsorbent using vapor, and then the desorbed vapor and VOCs were recovered and reused using a heat exchanger, and the time when the adsorbent was desorbed was detected using a temperature sensor.

[0005] However, even if the desorption time is detected, or the desorption time is exceeded and a warning is issued of the occurrence of an abnormality, if the operator does not take the appropriate steps to resolve the abnormality in a timely manner, the adsorption efficiency will decrease. Even if a timely response is made, the method of responding to the abnormality will differ depending on the operator's level of skill, resulting in the problem that the adsorption efficiency will not be maintained constant.

[0006] Furthermore, even if an artificial intelligence model is used, an artificial intelligence model that outputs an optimal adsorption / desorption cycle in accordance with a normal state may repeatedly output erroneous predicted values ​​in accordance with an abnormal state, resulting in erroneous learning, which may have an inaccurate effect on predictions of the adsorption / desorption cycle in the normal state. Summary of the Invention [Problem to be solved by the invention]

[0007] According to one aspect of the present disclosure, it is possible to provide an artificial intelligence-based ammonia absorption tower drive control device and a control method using the same, which, when an abnormal state occurs, performs an abnormal situation processing process in accordance with an exceptional adsorption period and an exceptional desorption period, regardless of a predicted adsorption / desorption period, thereby enabling environmentally friendly production of clean energy.

[0008] An ammonia absorber drive control device according to an embodiment of the present disclosure includes a sensor unit that measures a condition inside the absorber, a memory that stores one or more instructions, a processor that executes the one or more instructions stored in the memory, and a drive unit that drives the absorber according to an adsorption period and a desorption period that are set based on a condition inside the absorber, wherein the processor uses a learned artificial intelligence model to output a corresponding adsorption period and desorption period in accordance with sensing data from the sensor unit, and when the sensing data is within a preset optimal range, transmits a command to the drive unit to drive the absorber according to the output adsorption period and desorption period, and when the sensing data exceeds the optimal range, transmits a command to the drive unit to perform an abnormal situation processing process according to the preset exceptional adsorption period and exceptional desorption period.

[0009] According to an embodiment, the exceptional suction period and the exceptional desorption period may be a shortest suction period and a longest desorption period within a period range that can be set by the driving unit.

[0010] According to one embodiment, when the processor transmits an instruction to the driver to perform the abnormal situation handling process, the instruction may include at least one parameter of a minimum adsorption flow rate and a maximum desorption flow rate at which the driver can be operated, a minimum adsorption temperature and a maximum desorption temperature at which the driver can be operated, and a minimum adsorption pressure and a maximum desorption pressure at which the driver can be operated.

[0011] According to one embodiment, the processor may be configured to repeatedly self-diagnose at a preset period whether the sensing data measured by the sensor unit falls outside an optimal range or whether the patterns of previous and subsequent sensing data are different.

[0012] According to one embodiment, after driving the absorption tower according to the abnormal situation handling process, if the sensing data measured by the sensor unit is within an optimal range, a command to drive the absorption tower using the adsorption period and desorption period output from the trained artificial intelligence model can be transmitted to the driving unit.

[0013] According to one embodiment, the processor can be configured to calculate reduction data of cost and time when the drive unit is operating using the next adsorption cycle and the next desorption cycle when the previous adsorption cycle and the previous desorption cycle performed before driving the absorption tower according to the abnormal situation treatment process are different from the next adsorption cycle and the next desorption cycle for driving the absorption tower after the abnormal situation treatment process is completed.

[0014] According to one embodiment, the processor may transmit a command to the driving unit to maintain the previous adsorption cycle and the previous desorption cycle as they are if the calculated cost and time reduction data does not exceed a preset threshold.

[0015] According to one embodiment, the processor can remove abnormal state sensing data at each point in the process and transmit only normal state sensing data to the artificial intelligence model so that the artificial intelligence model can detect abnormal states.

[0016] According to an embodiment, the system may further include a user interface that provides an alarm to a user when the abnormal condition occurs and accepts input of information related to the abnormal situation handling process.

[0017] According to one embodiment, the sensor unit includes at least one sensor selected from the group consisting of a pressure sensor, a temperature sensor, a humidity sensor, a gas sensor, and an adsorbent sensor, and the processor can transmit a command to the drive unit to stop driving the absorption tower when an error in the at least one sensor is detected by self-diagnosis or a communication error between the at least one sensor and the processor is detected.

[0018] An ammonia absorber drive control method according to an embodiment of the present disclosure is executed by a computing device including a memory storing one or more instructions and a processor executing the one or more instructions stored in the memory. The method includes a step of collecting sensing data measuring a state in an absorber, and a step of driving the absorber along an adsorption period and a desorption period set based on the sensing data, and the step of driving the absorber may further include a step of outputting a corresponding adsorption period and a desorption period according to the sensing data using a trained artificial intelligence model, a step of transmitting a command to the drive unit to drive the absorber along the output adsorption period and desorption period when the sensing data is within a preset optimal range, and a step of transmitting a command to the drive unit to perform an abnormal situation processing process along the preset exceptional adsorption period and exceptional desorption period when the sensing data exceeds the optimal range. Effect of the Invention

[0019] According to one embodiment, the present disclosure provides the advantage that, when an abnormal condition occurs, an abnormal situation handling process is performed regardless of the predicted adsorption / desorption cycle to quickly restore conditions in the absorption tower to normal, thereby maximizing costs and adsorption efficiency.

[0020] In addition, the present disclosure provides the effect of maximizing adsorption efficiency by saving costs and time associated with changing the driving conditions, by verifying the driving conditions before operating the absorption tower in accordance with the predicted adsorption / desorption cycle output by the artificial intelligence model, and changing the driving conditions in accordance with the costs incurred when changing the adsorption / desorption cycle and the improved operating performance efficiency. [Brief description of the drawings]

[0021] [Figure 1] FIG. 1 shows a simplified conceptual diagram of an absorber tower according to one embodiment of the present disclosure. [Diagram 2] FIG. 2 is a block diagram simply illustrating a configuration of an ammonia absorber drive control device according to an embodiment of the present disclosure. [Diagram 3] FIG. 3 shows a flowchart of an ammonia absorber drive control method according to one embodiment of the present disclosure. [Figure 4] FIG. 4 shows a specific flowchart of an ammonia absorber drive control method according to one embodiment of the present disclosure. [Diagram 5] FIG. 5 shows a flowchart of the ammonia absorber drive control method according to one embodiment of the present disclosure when an abnormal state occurs. [Figure 6] FIG. 6 illustrates a computing device according to one embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0022] Hereinafter, embodiments of the present invention will be described in more detail with reference to the drawings. However, these embodiments are merely illustrative of the present invention and are not intended to limit the present invention.

[0023] The same reference numerals refer to the same components throughout this disclosure. This disclosure does not describe all elements of the embodiments, and general content in the technical field to which the disclosure belongs or content that is duplicated in the embodiments is omitted. The term "unit, module, member, block" used in the specification can be embodied in software or hardware, and depending on the embodiment, multiple "units, modules, members, blocks" may be embodied as one component, or one "unit, module, member, block" may include multiple components.

[0024] Throughout the specification, when a part is referred to as being "coupled" to another part, this includes not only direct coupling but also indirect coupling, including coupling via a wireless communication network.

[0025] Furthermore, unless otherwise specified, when a part "comprises" a certain component, it does not exclude other components but means that the part may further include other components.

[0026] Throughout the specification, a member is referred to as being "on" another member not only when the member is in contact with the other member, but also when there is another member between the two members.

[0027] The terms "first", "second", etc. are used only to distinguish one component from another component, and are not intended to limit the components.

[0028] The singular expression includes the plural expression unless the context clearly indicates otherwise.

[0029] The identification numbers in each step are used for ease of description and do not dictate the order of each step, and each step may be performed in a manner different from the stated order unless the context clearly dictates a particular order.

[0030] In this specification, the term "device according to the present disclosure" includes all of various devices capable of performing computation and providing a result to a user. For example, the device according to the present disclosure may include all of a computer, a server device, and a portable terminal, and may be in any form.

[0031] Here, the computer may include, for example, a notebook computer, a desktop computer, a laptop computer, a tablet PC, a slate PC, etc. equipped with a web browser.

[0032] The server device is a server that communicates with external devices and processes information, and may include an application server, a computing server, a database server, a file server, a game server, a mail server, a proxy server, a web server, and the like.

[0033] The portable terminal is, for example, a wireless communication device that ensures portability and mobility, and may include all kinds of handheld-based wireless communication devices such as PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), WiBro (Wireless Broadband Internet) terminals, and smartphones, as well as wearable devices such as watches, rings, bracelets, anklets, necklaces, glasses, contact lenses, or head-mounted devices (HMDs).

[0034] The artificial intelligence and related functions according to the present disclosure are operated by a processor and a memory. The processor may be composed of one or more processors. Here, the one or more processors may be general-purpose processors such as CPU, AP, DSP (Digital Signal Processor), graphics dedicated processors such as GPU, VPU (Vision Processing Unit), or artificial intelligence dedicated processors such as NPU. The one or more processors control to process input data according to predefined operation rules or artificial intelligence models stored in the memory. In addition, when the one or more processors are artificial intelligence dedicated processors, the artificial intelligence dedicated processors may be designed with a hardware structure specialized for processing a specific artificial intelligence model.

[0035] The predefined behavior rules or artificial intelligence models are characterized by being created through learning. Here, "created through learning" means that a predefined behavior rule or artificial intelligence model is created by learning a basic artificial intelligence model using a large amount of learning data by a learning algorithm, and the predefined behavior rule or artificial intelligence model is configured to perform a desired characteristic (or purpose). This learning may be performed on the device itself on which the artificial intelligence according to the present disclosure is executed, or may be performed by another server and / or system. Examples of learning algorithms include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0036] The artificial intelligence model may be composed of a plurality of neural network layers. Each of the plurality of neural network layers has a plurality of weight values, and performs a neural network operation through an operation between the operation result of a previous layer and the plurality of weights. The plurality of weights of the plurality of neural network layers may be optimized according to a learning result of the artificial intelligence model. For example, the plurality of weights may be updated so that a loss value or a cost value obtained from the artificial intelligence model during the learning process is reduced or minimized. The artificial neural network may include a deep neural network (DNN), such as, but not limited to, a convolutional neural network (CNN), a deep neural network (DNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRRDNN), or a deep Q-network.

[0037] According to an exemplary embodiment of the present disclosure, the processor may embody artificial intelligence. Artificial intelligence refers to a machine learning method based on an artificial neural network that imitates human biological neurons and allows a machine to learn. The methodology of artificial intelligence can be classified into supervised learning, in which input data and output data are provided together as training data according to a learning method, so that an answer (output data) to a problem (input data) is determined; unsupervised learning, in which only input data is provided without output data, so that an answer (output data) to a problem (input data) is not determined; and reinforcement learning, in which a reward is given from an external environment every time an action is taken in a current state, and learning is performed in a direction to maximize the reward. Artificial intelligence methodologies can also be categorized by architecture, which is the structure of the learning model. The architectures of widely used deep learning technologies can be categorized into convolutional neural networks (CNNs), recurrent neural networks (RNNs), Transformers, and generative adversarial networks (GANs), etc.

[0038] The device and system may include an artificial intelligence model. The artificial intelligence model may be one artificial intelligence model or may be embodied as multiple artificial intelligence models. The artificial intelligence model may be configured as a neural network (or an artificial neural network) and may include a statistical learning algorithm that mimics biological nerves in machine learning and cognitive science. A neural network may refer to a general model in which artificial neurons (nodes) forming a network by synaptic connections change the strength of synaptic connections through learning and have a problem-solving ability. Neurons in a neural network may include a combination of weights or biases. A neural network may include one or more layers composed of one or more neurons or nodes. For example, the device may include an input layer, a hidden layer, and an output layer. The neural network constituting the device may infer a result to be predicted from an arbitrary input by changing the weights of neurons through learning.

[0039] The processor may generate a neural network, train or learn the neural network, perform an operation based on received input data, generate an information signal based on the execution result, or retrain the neural network. The neural network model may include, but is not limited to, various types of models such as CNN (Convolution Neural Network) such as GoogleNet, AlexNet, and VGG Network, R-CNN (Region with Convolution Neural Network), RPN (Region Proposal Network), RNN (Recurrent Neural Network), S-DNN (Stacking-based deep Neural Network), S-SDNN (State-Space Dynamic Neural Network), Deconvolution Network, DBN (Deep Belief Network), RBM (Restricted Boltzman Machine), Fully Convolutional Network, LSTM (Long Short-Term Memory) Network, and Classification Network. The processor may include one or more processors for performing an operation according to the neural network model. For example, the neural network may include a deep neural network.

[0040] Neural networks include CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), perceptron, multilayer perceptron, FF (Feed Forward), RBF (Radial Basis Network), DFF (Deep Feed Forward), LSTM (Long Short Term Memory), GRU (Gated Recurrent Unit), AE (Auto Encoder), VAE (Variational Auto Encoder), DAE (Denoising Auto Encoder), SAE (Sparse Auto Encoder), MC (Markov Chain), HN (Hopfield Network), BM (Boltzmann Machine), RBM (Restricted Boltzmann Machine), DBN (Deep Belief Network), DCN (Deep Convolutional Network), DN (DeconvoluTional Network), DCIGN (Deep Convolutional Inverse Graphics Network), and GAN (Generative Adversarial Network), LSM (Liquid State Machine), ELM (Extreme Learning Machine), ESN (Echo State Network), DRN (Deep Residual Network), DNC (Differentiable Neural Computer), NTM (Neural Turning Machine), CN (Capsule Network), KN (Kohonen Network), and AN (Attention Network), but it will be understood by those of ordinary skill in the art that the neural network may include any neural network, without being limited thereto.

[0041] According to an exemplary embodiment of the present disclosure, the processor may support a variety of neural networks, including CNNs (Convolution Neural Networks) such as GoogleNet, AlexNet, and VGGNetwork, R-CNN (Region with Convolution Neural Network), RPN (Region Proposal Network), RNN (Recurrent Neural Network), S-DNN (Stacking-based deep Neural Network), S-SDNN (State-Space Dynamic Neural Network), Deconvolution Network, DBN (Deep Belief Network), RBM (Restricted Boltzman Machine), Fully Convolutional Network, LSTM (Long Short-Term Memory) Network, Classification Network, Generative Modeling, eXplainable AI, Continual AI, Representation Learning, AI for Material Design, BERT, SP-BERT, MRC / QA, Text Analysis, Dialog System, GPT-3, GPT-4 for Natural Language Processing, Visual Analytics, Visual Understanding, Video Synthesis, ResNet for Data Intelligence, Anomaly Detection, Prediction, Time-Series Various artificial intelligence structures and algorithms may be used, including but not limited to, forecasting, optimization, recommendation, data creation, etc. Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

[0042] FIG. 1 is a conceptual diagram showing a simplified configuration of an absorber 10 according to an embodiment of the present disclosure. Hereinafter, the AI-based ammonia absorber drive control device and a control method using the same will be described in detail with reference to FIGS. 2 to 5.

[0043] As shown in FIG. 1, the absorption tower 10 receives air (Feed) containing ammonia, volatile organic compounds (VOCs), etc., and applies vapor recovery methods such as recompression, condensation, and activated carbon adsorption between T1 and T4 to discharge air (Product) from which ammonia, volatile organic compounds (VOCs), etc. have been removed.

[0044] The absorption tower 10 can be divided into regions T1, T2, T3, and T4 from a position close to the inlet where the feed is input to the position of the outlet where the product is discharged, and a sensor unit 110 (see FIG. 2) can be provided in each divided region. The sensor unit 110 can include at least one of a pressure sensor that measures the pressure inside the absorption tower 10, a temperature sensor that measures the temperature inside the absorption tower 10, a humidity sensor that measures the humidity inside the absorption tower 10, a gas sensor that measures the gas components inside the absorption tower 10 to inspect the purity of the extracted hydrogen, and an adsorbent sensor.

[0045] Using the sensor unit 110, it is possible to check whether the pressure or temperature is being maintained within a certain range to maintain the adsorption efficiency, and the result value regarding the hydrogen extraction efficiency can be calculated using the humidity or hydrogen purity.

[0046] Furthermore, the adsorbent sensor of the sensor unit 110 can measure the state of the adsorbent in the absorption tower, evaluate the life and performance of the adsorbent, and obtain information that can predict the timing of replacing the adsorbent.

[0047] As shown in Figure 1, the more active the adsorption, the higher the temperature becomes. However, as the adsorption efficiency decreases and the time to replace the adsorbent approaches, the rate of temperature increase decreases. Therefore, the change in the state of the adsorbent can be detected from the amount of change in temperature.

[0048] In this case, it can be seen that the contact time with the feed becomes slower from T1 to T4, and the time to replace the adsorbent occurs later, and that the time to replace the adsorbent from T1 to T4 does not occur all at the same time.

[0049] Furthermore, when an abnormal state occurs, the range of temperature change in the absorption tower 10 may decrease, even though it is not time to replace the adsorbent, which may lead to the mistaken belief that it is time to replace the adsorbent.

[0050] Therefore, as shown in FIG. 2, an artificial intelligence-based ammonia absorber drive control device 100 according to an embodiment of the present disclosure may include a sensor unit 110 that measures a condition inside the absorber 10, a memory 16 that stores one or more instructions, a processor 14 that executes the one or more instructions stored in the memory 16, and a drive unit 130 that drives the absorber 10 according to an adsorption period and a desorption period that are set based on a condition inside the absorber 10.

[0051] The sensor units 110 can be disposed at T1 to T4 in the absorption tower 10, respectively.

[0052] As one embodiment, the inside of the absorption tower 10 can be separated and partitioned into a plurality of beds, and the sensor unit 110 can be provided in each of the separated and partitioned beds.

[0053] In another embodiment, a plurality of absorption towers 10 may be connected together, and each absorption tower 10 may be provided with a sensor unit 110 .

[0054] In other words, a sensor unit 110 can be provided for each space separated into at least one absorption tower 10, which is separated into separate spaces, such as by dividing one absorption tower 10 into multiple beds or by connecting multiple absorption towers 10 together.

[0055] The processor 14 and the memory 16 may be located together or respectively in one or more computing devices 120. The computing devices 120 may be used to control the operation of the absorber 10 so as to maintain the efficiency of the absorber 10 at or above a certain level.

[0056] Artificial intelligence model 160 may be co-located on the computing device 120 or may be linked via another server or the cloud.

[0057] In addition, the AI-based ammonia absorber drive control device 100 according to an embodiment of the present disclosure may further include a user interface 140 for receiving a user's input regarding the operation of the absorber 10.

[0058] The user interface 140 can provide an alarm to the user in the event of an abnormal condition and can accept input of information related to the abnormal situation handling process.

[0059] In addition, the AI-based ammonia absorber drive control device 100 according to an embodiment of the present disclosure may further include a display unit 150 that provides an alarm regarding the operation of the absorber 10 or monitors the operating status of the absorber 10 in real time and displays the monitoring results.

[0060] In one embodiment, the processor 14 uses the learned artificial intelligence model 160 to output corresponding adsorption periods and desorption periods according to the sensing data of the sensor unit 110, and if the sensing data is within a preset optimal range, transmits a command to the drive unit 130 to drive the absorption tower according to the output adsorption period and desorption period, and if the sensing data exceeds the optimal range, transmits a command to the drive unit 130 to perform an abnormal situation handling process according to the preset exception adsorption period and exception desorption period.

[0061] The AI-based ammonia absorption tower drive control device 100 identifies an abnormal state based on data collected by the sensor unit 110, and can detect the abnormal state in real time when an abnormal state occurs by using the AI ​​model 160 that has learned normal states such as pressure, temperature, humidity, and concentration of the gas in the absorption tower 10 according to each point in the adsorption process.

[0062] The processor 14 of the AI-based ammonia absorption tower drive control device 100 repeatedly performs self-diagnosis at a preset period to determine whether the sensing data measured by the sensor unit 110 is outside an optimal range or whether the patterns of the previous sensing data and the subsequent sensing data are different, and can detect an abnormal state in real time when an abnormal state occurs using the AI ​​model 160.

[0063] In addition, when the processor 14 proceeds with self-diagnosis, it not only detects abnormal conditions within the absorption tower 10, but also detects the occurrence of a sensor error in at least one sensor of the sensor unit 110, or a communication error between the sensor unit 110 and the processor 14, and upon detecting the occurrence of a sensor error or a communication error, it can transmit a command to the drive unit 130 to stop operating the absorption tower 10.

[0064] In one embodiment, when the normal state is identified, the processor 14 may perform the adsorption process at an optimal efficiency using the adsorption period and the desorption period output by the artificial intelligence model 160. In this case, the optimal efficiency means a state in which the adsorption efficiency can be maintained at a specific level or higher while reducing costs by delaying the replacement timing of the adsorbent as much as possible.

[0065] In one embodiment, when an abnormal state is identified, the processor 14 does not perform the adsorption process in the adsorption and desorption cycles output by the artificial intelligence model 160, but instead operates a preset abnormal situation handling process to command the operation of the absorber 10 in the adsorption / desorption cycle that will allow the system to return to a normal state most quickly.

[0066] Here, the adsorption / desorption period that returns to the normal state the fastest is a preset exceptional adsorption period and exceptional desorption period, and the exceptional adsorption period and exceptional desorption period may be the shortest adsorption period and the longest desorption period within a period range that can be set by the driving unit 130.

[0067] Alternatively, when the processor 14 transmits an instruction to the driver 130 to perform the abnormal situation handling process, the instruction may include parameters according to at least one of the minimum adsorption flow rate and maximum desorption flow rate, the minimum adsorption temperature and maximum desorption temperature, or the minimum adsorption pressure and maximum desorption pressure that the driver 130 can drive.

[0068] Specifically, the processor 14 can transmit exception parameters corresponding to abnormal situations other than the exception adsorption cycle and the exception desorption cycle to the driver 130. The minimum adsorption flow rate and maximum desorption flow rate at which the driver 130 can be driven can be transmitted to the driver 130, and the minimum adsorption flow rate that can be handled while maintaining optimal adsorption efficiency before the adsorbent reaches the replacement cycle can be measured or calculated, and the maximum desorption flow rate that can be handled while maintaining optimal desorption efficiency before the adsorbent reaches the replacement cycle can be measured or calculated.

[0069] Alternatively, the processor 14 can transmit the minimum adsorption temperature and maximum desorption temperature, or the minimum adsorption pressure and maximum desorption pressure, at which the drive unit 130 can be operated, and measure or calculate the minimum adsorption temperature or pressure that occurs when the adsorbent maintains the optimal adsorption efficiency before reaching the replacement cycle, and measure or calculate the maximum desorption temperature or pressure that occurs when the adsorbent maintains the optimal desorption efficiency before reaching the replacement cycle.

[0070] At this time, the adsorption period and desorption period output by the artificial intelligence model 160 can be separately stored so that they can be used when the normal state is restored later.

[0071] In one embodiment, after identifying an abnormal state, the absorber 10 is driven in accordance with the abnormal situation processing process, and then, if the sensing data measured by the sensor unit 110 is again within the optimal range, the absorber 10 can be driven using the adsorption period and desorption period output from the trained artificial intelligence model 160.

[0072] At this time, the artificial intelligence model 160 can be operated again to output the adsorption period and desorption period, or the adsorption period and desorption period output by the artificial intelligence model 160 that were previously stored separately can be called up and used.

[0073] Therefore, the AI-based ammonia absorber drive control device 100 can distinguish between normal and abnormal states in advance and control the operation of the absorber 10 to perform a separate adsorption process accordingly.

[0074] Meanwhile, the processor 14 of the artificial intelligence-based ammonia absorber drive control device 100 may be configured to calculate cost and time reduction data when the drive unit is driven using the next adsorption cycle and the next desorption cycle, when the previous adsorption cycle and the previous desorption cycle performed before driving the absorber 10 according to the abnormal situation processing process are different from the next adsorption cycle and the next desorption cycle for driving the absorber after the abnormal situation processing process is completed.

[0075] Specifically, when there exists an adsorption / desorption period (P1) in which the absorption tower 10 was driven before the abnormal state was identified, an exceptional adsorption / desorption period (P2), and an exceptional adsorption / desorption period (P2) output again by the artificial intelligence model 160 after the normal state is restored, the processor 14 does not immediately perform the adsorption process using P2, but can perform the adsorption process after performing a final verification of the adsorption processes using P1 and P2.

[0076] For example, when changing the operating condition of the absorber 10 from P1 to P2, the cost incurred and the improved operating performance efficiency are calculated and compared, and then the adsorption / desorption cycle can be changed.

[0077] In other words, the calculated planned cost and efficiency of changing the adsorption / desorption cycle are compared with the current cost and efficiency of the adsorption / desorption cycle, and the absorption tower 10 can be controlled to change the adsorption / desorption cycle only if there is an effect of reducing costs and increasing efficiency equal to or greater than a preset threshold value.

[0078] It is possible to control the absorber 10 in a manner that reflects the overall operating status of the absorber 10 by taking into consideration all of the losses associated with frequent changes in the adsorption / desorption cycle and the process delays in the cycle change process.

[0079] Therefore, the processor 14 of the artificial intelligence-based ammonia absorption tower drive control device 100 can transmit a command to the drive unit 130 to maintain the previous adsorption cycle and the previous desorption cycle as they are if the calculated cost and time reduction data does not exceed a preset threshold.

[0080] The processor 14 of the AI-based ammonia absorber drive control device 100 can transmit drive data of the absorber 10 as learning data to the AI ​​model 160. In order for the AI ​​model 160 to detect an abnormal state, it is possible to remove sensing data of an abnormal state at each time point of the process and transmit only sensing data of a normal state to the AI ​​model.

[0081] Meanwhile, as shown in FIG. 3, a control method using an AI-based ammonia absorber drive control device 100 may include a step (S310) of collecting sensing data measuring a state inside the absorber 10, and a step (S320) of driving the absorber 10 according to an adsorption period and a desorption period set based on the sensing data.

[0082] The step of driving the absorber 10 may further include the steps of: outputting corresponding adsorption and desorption periods according to the sensing data using the trained artificial intelligence model 160; transmitting, when the sensing data is within a preset optimal range, a command to the driving unit 130 to drive the absorber 10 according to the output adsorption and desorption periods; and transmitting, when the sensing data is outside the optimal range, a command to the driving unit 130 to perform an abnormal situation handling process according to the preset exception adsorption and desorption periods.

[0083] 4, in a data collection step (S410), sensing data in the initial absorption tower 10 can be collected via the sensor unit 110. Then, in a data preprocessing step (S420), the sensing data can be subjected to data preprocessing such as data normalization, missing value processing, and removal of abnormal values ​​measured under abnormal conditions.

[0084] In the artificial intelligence model learning step (S430), the artificial intelligence model 160 is trained using the preprocessed data to generate a model for predicting optimal operating conditions of the absorption tower 10. Using a machine learning or deep learning algorithm, the artificial intelligence model 160 can be trained to output an optimal adsorption / desorption cycle corresponding to the sensing data, and in particular, the artificial intelligence model 160 can be trained to output an optimal adsorption / desorption cycle taking into consideration the replacement cycle and cost of the adsorbent, the optimal adsorption efficiency, the time for each process for maintaining the optimal adsorption efficiency, and the purity of the gas.

[0085] When the learning is completed, in an adsorption / desorption cycle output step (S440) using the learned artificial intelligence model 160, an optimal adsorption / desorption cycle corresponding to the current state inside the absorber 10 can be output.

[0086] Instead of immediately performing the adsorption / desorption process in the predicted adsorption / desorption cycle, an abnormal state detection step (S450) can be performed, and if no abnormal state is detected, a drive condition verification step (S460) can be performed.

[0087] In the abnormal state detection step (S450), the artificial intelligence model 160 is used to detect deviations from the expected temperature, humidity, or pressure range based on the current amount of adsorbent and the process progress status in the absorption tower 10, and the difference between the expected predicted value and the current actual measured value is determined. If the difference is equal to or greater than a threshold, it is identified as an abnormal state and an alarm is provided on the display unit 150.

[0088] The difference between the predicted value and the current measured value is determined, and if the difference is less than a threshold and does not fall outside the normal state range, a drive condition verification step (S460) can determine whether to maintain or change the existing adsorption / desorption cycle based on the amount of adsorbent, the amount of energy, and the application time for each step required for gas production.

[0089] Finally, with the verification of the driving conditions being completed, the driving step (S470) of the absorber 10 can be performed at the determined adsorption / desorption cycle.

[0090] Alternatively, as shown in Fig. 5, when an abnormal state is identified in the abnormal state detection step (S510), an abnormal situation processing process execution step (S520) can be performed according to the exception adsorption / desorption cycle. The abnormal state detection step (S510) is the same step as S410.

[0091] In the abnormal situation processing process execution step (S520), the adsorption and desorption process is performed according to the preset exceptional adsorption period and exceptional desorption period, so that the state inside the absorber 10 can be restored to the normal state as quickly as possible.

[0092] Thereafter, in a normal state return detection step (S530), a self-diagnosis is performed by performing an adsorption / desorption process according to the exceptional adsorption cycle and the exceptional desorption cycle, and when the sensing data measured by the sensor unit 110 falls within the normal state range, a return to the normal state is detected, and an adsorption / desorption cycle output step (S540) can be performed using the learned artificial intelligence model 160. The adsorption / desorption cycle output step (S540) using the learned artificial intelligence model 160 is the same step as S440. After S540, S450, S460, and S470 can be performed in succession.

[0093] In addition, the processor 14 of the AI-based ammonia absorber drive control device 100 drives the absorber 10 according to the output adsorption cycle and desorption cycle, and then, when the actual process result measured by the sensor unit 110 exceeds a preset range of the predicted process result, the processor 14 can transmit a command to the drive unit 130 to perform an abnormal situation processing process according to the exceptional adsorption cycle and the exceptional desorption cycle.

[0094] Alternatively, the processor 14 of the AI-based ammonia absorber drive control device 100 can drive the absorber 10 according to the output adsorption cycle and desorption cycle, and then transmit a re-learning command to the AI ​​model 160 when the actual process result measured by the sensor unit 110 exceeds a preset range of the predicted process result.

[0095] If the sensing data of the sensor unit 110 does not deviate from the normal range but the adsorption efficiency is not as expected, the abnormal situation processing process can be activated to reset the state within the absorption tower 10, or re-learning can be performed to eliminate the possibility of erroneous learning of the artificial intelligence model 160.

[0096] In another embodiment, the processor 14 of the artificial intelligence-based ammonia absorption tower drive control device 100 can determine whether the abnormal condition is temporary or permanent, and apply different abnormal situation handling processes depending on the abnormal condition.

[0097] In the case of a temporary erroneous measurement by the sensor unit 110 or a communication network failure due to an external factor, since this is a temporary abnormal state, the adsorption / desorption process can be performed using a preset exceptional adsorption / desorption cycle.

[0098] If the abnormal condition is due to a failure of the sensor unit 110 or a communication error between at least one of the sensor units 110 and the processor 14 due to a failure of the communication unit, the operation of the absorption tower 10 can be stopped since the abnormal condition is a permanent abnormal condition that will not be resolved unless repaired externally.

[0099] Alternatively, in the case of a permanent abnormal state, the average value of the output adsorption / desorption cycles stored in the memory 16 can be calculated, and the adsorption / desorption process can be performed according to the calculated adsorption / desorption cycle. In order to prevent costs and losses while waiting for the abnormal state to be resolved, the absorber 10 can be operated at any efficiency.

[0100] In another embodiment, an abnormal state is an exceptional pattern that deviates from a normal state, and the artificial intelligence model 160 can learn a pattern based on parameters measured in a normal state, for example, the temperature change in the absorption tower 10, and detect an abnormal state if the pattern changes.

[0101] 1, when a temperature pattern different from that in a normal state is shown in an abnormal state, the artificial intelligence model 160 can detect the difference from the learned pattern and proceed with the abnormal situation processing process. In this case, the pattern can include a temperature pattern, a pressure pattern, a humidity pattern, a gas concentration pattern, etc.

[0102] As a result, the desorption process is performed at an appropriate time so that the efficiency of the absorber 10 can be maintained at a certain level or above and the adsorbent can be used continuously, and the absorber 10 can be operated stably not only under normal conditions but also under abnormal conditions.

[0103] In addition, by operating the absorber 10 while taking into consideration the cost and adsorption efficiency through simulations before and after control of the adsorption and desorption cycles, it is possible to reduce costs and maintain optimal adsorption efficiency for a long period of time.

[0104] Meanwhile, as shown in Fig. 6, the AI-based ammonia absorption tower drive control device 100 according to the present disclosure may be embodied in the form of a storage medium storing computer-executable instructions. The instructions may be stored in the form of program code, and when executed by a processor, may generate a program module to perform the operations of the disclosed embodiments. The storage medium may be embodied as a computer-readable storage medium.

[0105] The artificial intelligence-based ammonia absorption tower drive control device according to the present disclosure may correspond to a computing device 12. The computing device 12 may include at least one processor 14, a computer-readable storage medium 16 including a program 20, and a communication bus 18. The computing device 12 may also include one or more input / output interfaces 22 that provide an interface for an input / output device 24, and one or more network communication interfaces 26.

[0106] The computing device 120 of the present disclosure may correspond to the computing device 12. The user interface 140 and the display unit 150 of the present disclosure may correspond to one or more I / O interfaces 22 that provide an interface for the I / O device 24. Therefore, each component of the AI-based ammonia absorber drive control device 100 according to the present disclosure may be included in the single computing device 12 described above, or may be embodied in a separate device.

[0107] The disclosed embodiments have been described above with reference to the accompanying drawings. Those skilled in the art will understand that the present disclosure can be implemented in a form different from the disclosed embodiments without changing the technical idea or essential features of the present disclosure. It should be understood that the disclosed embodiments are illustrative and not limiting. [Explanation of symbols]

[0108] 100 Ammonia absorption tower drive control device 110 Sensor section 120 Computing Devices 14 processors 16 Memory 130 Drive unit 140 User Interface 150 Display unit 160 Artificial Intelligence Models

Claims

1. A sensor unit for measuring a state inside the absorption tower; a memory storing one or more instructions; a processor for executing the one or more instructions stored in the memory; a drive unit that drives the absorption tower according to an adsorption period and a desorption period that are set based on a state inside the absorption tower, The processor, Using the learned artificial intelligence model, outputting a corresponding adsorption period and a corresponding desorption period in accordance with the sensing data of the sensor unit; If the sensing data is within a preset optimum range, a command to drive the absorption tower according to the output adsorption cycle and desorption cycle is transmitted to the driving unit; If the sensing data exceeds the optimal range, a command is sent to the driving unit to perform an abnormal situation processing process according to a preset exception adsorption period and an exception desorption period. Ammonia absorber drive control device.

2. The exceptional adsorption period and the exceptional desorption period are a shortest adsorption period and a longest desorption period within a period range that can be set by the drive unit. The ammonia absorber drive control device according to claim 1.

3. When the processor transmits a command to the driving unit to perform the abnormal situation handling process, the command The minimum adsorption flow rate and the maximum desorption flow rate that can be driven by the driving unit; The minimum adsorption temperature and maximum desorption temperature at which the driving unit can be driven, and The parameter includes at least one of a minimum adsorption pressure and a maximum desorption pressure at which the driving unit can be driven. The ammonia absorber drive control device according to claim 2.

4. The processor, The device is configured to repeatedly perform self-diagnosis at a preset period to check whether the sensing data measured by the sensor unit falls outside an optimum range or whether the pattern of the previous sensing data differs from that of the subsequent sensing data. The ammonia absorber drive control device according to claim 2.

5. After driving the absorption tower according to the abnormal situation processing process, if the sensing data measured by the sensor unit is within an optimal range, a command to drive the absorption tower using the adsorption period and desorption period output from the trained artificial intelligence model is transmitted to the driving unit. The ammonia absorber drive control device according to claim 4.

6. The processor, When the previous adsorption cycle and the previous desorption cycle performed before driving the absorption tower according to the abnormal situation treatment process are different from the next adsorption cycle and the next desorption cycle in which the absorption tower is driven after the abnormal situation treatment process is completed, The drive unit is configured to calculate reduction data of cost and time during operation using the next adsorption cycle and the next desorption cycle. The ammonia absorber drive control device according to claim 5.

7. The processor, If the calculated cost and time reduction data does not exceed a preset threshold, a command to maintain the previous adsorption cycle and the previous desorption cycle is transmitted to the driving unit. The ammonia absorber drive control device according to claim 6.

8. The processor removes abnormal sensing data at each time point of the process and transmits only normal sensing data to the artificial intelligence model so that the artificial intelligence model detects the abnormal state. The ammonia absorber drive control device according to claim 1.

9. a user interface for providing an alarm to a user when the abnormal condition occurs and for receiving information input regarding the abnormal situation handling process; The ammonia absorber drive control device according to claim 8.

10. the sensor unit includes at least one sensor selected from the group consisting of a pressure sensor, a temperature sensor, a humidity sensor, a gas sensor, and an adsorbent sensor; When an error in the at least one sensor is detected by self-diagnosis or a communication error between the at least one sensor and the processor is detected, the processor transmits a command to the drive unit to stop driving the absorption tower. The ammonia absorber drive control device according to claim 4.

11. 1. A method for controlling an ammonia absorber, the method being executed by a computing device including a memory storing one or more instructions and a processor executing the one or more instructions stored in the memory, the method comprising: A step of collecting sensing data measuring a state inside the absorption tower; and driving the absorption tower according to an adsorption cycle and a desorption cycle that are set based on the sensing data, The step of driving the absorption tower includes: outputting corresponding adsorption and desorption periods according to the sensing data using the trained artificial intelligence model; transmitting a command to a drive unit of an absorber to drive the absorber according to the outputted adsorption cycle and desorption cycle when the sensing data is within a preset optimum range; and if the sensing data exceeds the optimal range, transmitting a command to a driving unit of the absorption tower to perform an abnormal situation processing process according to a preset exception adsorption period and an exception desorption period. Method for controlling the operation of an ammonia absorber.

Images