KR102629022B1 - Scheduling method based on reinforcement learning - Google Patents

Abstract

According to the present disclosure, a method for generating a schedule of a time-series task is performed. Specifically, according to the present disclosure, a computing device identifies a plurality of sub-actions constituting an individual action, generates a feature set for the plurality of sub-actions, and bases the feature set on each of the sub-actions. An action matrix is created, and an artificial neural network model is used to create a schedule for performing the task based on the action matrix.

Description

Scheduling method based on reinforcement learning {SCHEDULING METHOD BASED ON REINFORCEMENT LEARNING}

This disclosure relates to a scheduling method based on reinforcement learning. Specifically, a schedule for generating an action matrix for a plurality of actions constituting a time-series task and performing tasks based on the action matrix using an artificial neural network model. It is about how to create .

Time-series tasks, many of which can be automated, are often observed in modern society. For example, assembling industrial products using multiple robot arms, performing various processes on raw materials using processing equipment, and allocating operations to be executed on each core of the CPU in real time can be considered time-series tasks. You can.

For these time-series tasks, many optimization methods have been researched to perform tasks quickly and efficiently to reduce cost and time. Recently, due to remarkable developments in the field of artificial intelligence, especially machine learning through deep learning, methods for performing such optimization using deep learning are being studied.

However, in many cases, the process at each point in time-series work consists of multiple smaller processes, which creates difficulties in developing reinforcement learning models that solve these complex problems.

In addition, even if a reinforcement learning model is used to automate these time-series tasks, if the type of task or the detailed specifications that make up the task change, there is a need to re-update and deploy the reinforcement learning model currently in use. . In relation to this, there are many difficulties in updating reinforcement learning models for workers in the field who lack knowledge of artificial neural network model design, so a process or system is needed to automatically update reinforcement learning models and distribute them to users through easy operation. .

Therefore, a reinforcement learning model that can create an optimized work plan for time series work in which a process consisting of multiple smaller processes exists, and a platform that automatically updates and distributes the reinforcement learning model in accordance with changes in the specifications of the time series work. There is a need in the art for a method to provide it.

Korean registered patent KR 2035389 discloses a process control method and system through neural network learning based on history data.

The present disclosure was developed in response to the above-described background technology, and its purpose is to utilize an artificial neural network model to generate a schedule for performing tasks based on an action matrix. Specifically, the present disclosure identifies a plurality of sub-actions constituting an individual action, generates a feature set and action matrix for the sub-actions, and works based on the action matrix using an artificial neural network model. The purpose is to create a schedule to perform.

Meanwhile, the technical problem to be achieved by the present disclosure is not limited to the technical problems mentioned above, and may include various technical problems within the scope of what is apparent to those skilled in the art from the contents described below.

According to an embodiment of the present disclosure for realizing the above-described problem, a method performed by a computing device to generate a schedule of a time-series task including a plurality of actions is disclosed. . The method includes identifying a plurality of sub-actions constituting an individual action; generating a feature set for the plurality of sub-actions; It includes generating an action matrix based on a feature set for each of the sub-actions and generating a schedule for performing the task based on the action matrix using an artificial neural network model. can do.

In one embodiment, generating a feature set for the plurality of sub-actions includes: identifying a selectable value for each of the plurality of sub-actions and identifying a selectable value for each of the sub-actions. It may include generating a feature set corresponding to each sub-action based on inputting information into an encoder-type artificial neural network.

In one embodiment, the action matrix is a multidimensional matrix constructed based on a feature set for each sub-action, and the dimension of the action matrix may be equal to the number of the plurality of sub-actions.

In one embodiment, the step of generating a schedule for performing the task based on the action matrix using the artificial neural network model: Corresponding to each time point based on the action matrix using the artificial neural network model It may include determining an action to be performed and creating a schedule for performing the task based on the action determined for each time point.

In one embodiment, an artificial neural network model that generates a schedule of time-series tasks performed by a computing device according to an embodiment of the present disclosure for realizing the task described above by utilizing the artificial neural network model. A method of providing through an artificial intelligence platform is disclosed. The method includes determining whether an update to the artificial neural network model is necessary and providing update information to the artificial neural network model through the artificial intelligence platform, wherein the artificial neural network model is configured to perform individual actions. A model that generates an action matrix based on a set of features for a plurality of sub-actions constituting the, and generates a schedule of the time-series work based on the action matrix. It can be.

In one embodiment, determining whether an update to the artificial neural network model is necessary may include determining whether additional sub-actions other than the plurality of sub-actions need to be considered.

In one embodiment, providing update information for the artificial neural network model through the artificial intelligence platform includes: generating an updated feature set in which features of the additional sub-action are reflected; It may include generating an updated action matrix based on the updated feature set and providing the updated action matrix through the artificial intelligence platform.

According to an embodiment of the present disclosure for realizing the above-described problem, a method performed by a computing device to generate a schedule of a time-series task is disclosed. The method includes identifying a plurality of sub-actions constituting an individual action; generating a feature set for the plurality of sub-actions; It includes generating an action matrix based on a feature set for each of the sub-actions and generating a schedule for performing the task based on the action matrix using an artificial neural network model. can do.

In one embodiment, generating a feature set for the plurality of sub-actions includes: identifying a selectable value for each of the plurality of sub-actions and identifying a selectable value for each of the sub-actions. It may include generating a feature set corresponding to each sub-action based on inputting information into an encoder-type artificial neural network.

In one embodiment, the action matrix is a multidimensional matrix constructed based on a feature set for each sub-action, and the dimension of the action matrix may be equal to the number of the plurality of sub-actions.

In one embodiment, the step of generating a schedule for performing the task based on the action matrix using the artificial neural network model: Corresponding to each time point based on the action matrix using the artificial neural network model It may include determining an action to be performed and creating a schedule for performing the task based on the action determined for each time point.

In one embodiment, the step of determining an action corresponding to each time point based on the action matrix using the artificial neural network model includes: Referring to the updated action matrix, the output value of the artificial neural network model for a specific time point is It may include generating a step and identifying an action corresponding to the output value.

In one embodiment, the artificial neural network model may be configured to select and output one of the elements of the action matrix updated at each time point, and the output value at each time point may be configured to be the input value at the next time point.

In one embodiment, the updated action matrix is an action in which elements related to already performed actions among elements present in the action matrix and elements related to actions that cannot be performed at the specific point in time among elements present in the action matrix are masked. It could be a matrix.

In one embodiment, the method includes generating a reward for a reinforcement learning agent for each time point based on at least one of cost or time required to perform an action corresponding to each time point; It may further include generating state information of the reinforcement learning agent for each time point and performing learning of the reinforcement learning agent based on the action, reward, and state information for each time point.

In one embodiment, the reinforcement learning agent may be trained to minimize at least one of cost or time required to perform the task.

According to an embodiment of the present disclosure for realizing the above-described object, a computer program stored in a computer-readable storage medium that causes a computing device to perform operations for generating a schedule of a time-series task is disclosed. The operations include: identifying a plurality of sub-actions constituting an individual action; An operation of generating a feature set for the plurality of sub-actions; Generating an action matrix based on a set of features for each of the sub-actions and generating a schedule for performing the task based on the action matrix using an artificial neural network model. can do.

According to an embodiment of the present disclosure for realizing the above-described problem, a computing device for performing an operation of generating a schedule of a time-series task is disclosed. The computing device includes a processor including one or more cores and a memory, wherein the processor identifies a plurality of sub-actions constituting an individual action, generates a feature set for the plurality of sub-actions, and generates a feature set for the plurality of sub-actions. An action matrix can be created based on a set of features for each action, and an artificial neural network model can be used to create a schedule for performing the task based on the action matrix.

This disclosure has the effect of allowing scheduling using reinforcement learning to be performed more quickly and accurately. In particular, in this disclosure, an action matrix is generated based on the results of inputting each sub-action into an encoder-type artificial neural network, and the actions of the reinforcement learning agent are output based on the action matrix, thereby enabling faster and more accurate scheduling using reinforcement learning. It can be done.

1 is a block diagram of a computing device for generating a schedule of a time-series task according to an embodiment of the present disclosure.
Figure 2 is a schematic diagram showing a network function according to an embodiment of the present disclosure.
Figure 3 is a conceptual diagram showing a reinforcement learning algorithm according to an embodiment of the present disclosure.
Figure 4 is a flowchart showing a process for creating a schedule for a time-series task according to an embodiment of the present disclosure.
FIG. 5 is a conceptual diagram illustrating a process for generating an action matrix when the number of sub-actions is 2 and a process for generating a schedule based on the action matrix using an artificial neural network model according to an embodiment of the present disclosure.
6 is a brief, general schematic diagram of an example computing environment in which embodiments of the present disclosure may be implemented.

The present disclosure identifies a plurality of sub-actions constituting an individual action, generates a feature set for each of the sub-actions, and generates an action matrix based on the feature set for each of the sub-actions. A method of generating a schedule for performing the task based on the action matrix using an artificial neural network model is disclosed.

Various embodiments are now described with reference to the drawings. In this specification, various descriptions are presented to provide an understanding of the disclosure. However, it is clear that these embodiments may be practiced without these specific descriptions.

As used herein, the terms “component,” “module,” “system,” and the like refer to a computer-related entity, hardware, firmware, software, a combination of software and hardware, or an implementation of software. For example, a component may be, but is not limited to, a process running on a processor, a processor, an object, a thread of execution, a program, and/or a computer. For example, both an application running on a computing device and the computing device can be a component. One or more components may reside within a processor and/or thread of execution. A component may be localized within one computer. In the present disclosure, one or more components may be distributed between two or more computers, or execution environments. Additionally, these components can execute from various computer-readable media having various data structures stored thereon. Components can transmit signals, for example, with one or more data packets (e.g., data and/or signals from one component interacting with other components in a local system, a distributed system, to other systems and over a network such as the Internet). Depending on the data being transmitted, they may communicate through local and/or remote processes.

Additionally, the term “or” is intended to mean an inclusive “or” and not an exclusive “or.” That is, unless otherwise specified or clear from context, “X utilizes A or B” is intended to mean one of the natural implicit substitutions. That is, either X uses A; X uses B; Or, if X uses both A and B, “X uses A or B” can apply to either of these cases. Additionally, the term “and/or” as used herein should be understood to refer to and include all possible combinations of one or more of the related listed items.

Additionally, the terms “comprise” and/or “comprising” should be understood to mean that the corresponding feature and/or element is present. However, the terms “comprise” and/or “comprising” should be understood as not excluding the presence or addition of one or more other features, elements and/or groups thereof. Additionally, unless otherwise specified or the context is clear to indicate a singular form, the singular terms herein and in the claims should generally be construed to mean “one or more.”

And, the term “at least one of A or B” should be interpreted to mean “when it contains only A,” “when it contains only B,” or “when it is a combination of A and B.”

Those skilled in the art will additionally understand that the various illustrative logical blocks, configurations, modules, circuits, means, logics, and algorithm steps described in connection with the embodiments disclosed herein may be implemented in electronic hardware, computer software, or both. It should be recognized that it can be implemented with combinations of. To clearly illustrate the interchangeability of hardware and software, various illustrative components, blocks, configurations, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented in hardware or software will depend on the specific application and design constraints imposed on the overall system. A skilled technician can implement the described functionality in a variety of ways for each specific application. However, such implementation decisions should not be construed as causing a departure from the scope of the present disclosure.

The description of the presented embodiments is provided to enable anyone skilled in the art to use or practice the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art. The general principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Therefore, the present disclosure is not limited to the embodiments presented herein. This disclosure is to be interpreted in the broadest scope consistent with the principles and novel features presented herein.

In the present disclosure, ‘time-series work’ may mean a set of ‘actions’ performed at a plurality of different points in time. Additionally, in the present disclosure, ‘sub-action’ may mean one or more sub-concepts constituting one action. For example, when manufacturing multiple dolls on one line in a doll factory is viewed as a time-series operation, manufacturing doll A at a specific point in time may be an action, and attaching eyes to doll A may be a sub-action. there is.

1 is a block diagram of a computing device for generating a schedule of a time-series task according to an embodiment of the present disclosure.

The configuration of the computing device 100 shown in FIG. 1 is only a simplified example. In one embodiment of the present disclosure, the computing device 100 may include different configurations for performing the computing environment of the computing device 100, and only some of the disclosed configurations may configure the computing device 100.

The computing device 100 may include a processor 110, a memory 130, and a network unit 150.

The processor 110 may be composed of one or more cores, and may include a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU) of a computing device. unit) may include a processor for data analysis and deep learning. The processor 110 may read a computer program stored in the memory 130 and perform data processing for machine learning according to an embodiment of the present disclosure. According to an embodiment of the present disclosure, the processor 110 may perform an operation for learning a neural network. The processor 110 is used for learning neural networks, such as processing input data for learning in deep learning (DL), extracting features from input data, calculating errors, and updating the weights of the neural network using backpropagation. Calculations can be performed.

At least one of the CPU, GPGPU, and TPU of the processor 110 may process learning of the network function. For example, CPU and GPGPU can work together to process learning of network functions and data classification using network functions. Additionally, in one embodiment of the present disclosure, the processors of a plurality of computing devices can be used together to process learning of network functions and data classification using network functions. Additionally, a computer program executed in a computing device according to an embodiment of the present disclosure may be a CPU, GPGPU, or TPU executable program.

The processor 110 may identify a time-series task to be performed and a plurality of sub-actions that constitute individual actions for performing the task. For example, if the task that the processor 110 must perform is a task of performing welding on a plurality of welding points included in one product using a plurality of welding equipment, which equipment is assigned to one welding point? Performing welding corresponds to 'action', and 'determining the equipment to perform welding' and 'determining the welding point' correspond to sub-actions.

The processor 110 may generate a feature set for a plurality of sub-actions. For example, if an individual action of a task consists of two subactions, a first subaction and a second subaction, a feature set for the first subaction may be created, and a feature set for the second subaction This can be created. A specific method for generating a feature set will be described later with reference to FIG. 3.

The processor 110 may generate an action matrix based on a feature set for each sub-action. When an individual action of a task consists of two subactions, a first subaction and a second subaction, the action matrix is a two-dimensional matrix obtained by matrix calculation of the feature set for the first subaction and the feature set for the second subaction. It can mean. In another embodiment, when individual actions of a task consist of a third subaction, a fourth subaction, and a fifth subaction, the action matrix may be a three-dimensional matrix. Accordingly, in the present disclosure, the action matrix may be a multi-dimensional matrix, and the dimension of the action matrix may be equal to the number of sub-actions constituting an individual action.

The processor 110 may utilize an artificial neural network model to create a schedule for performing tasks based on the action matrix. Specifically, the processor 110 may utilize an artificial neural network model to determine the action corresponding to that time point by referring to the action matrix at each time point.

In this disclosure, the artificial neural network utilized when creating a schedule to perform a task is an artificial neural network for processing time series input, such as a Recurrent Neural Network (RNN) or Long-Short Term Memory (LSTM). You can. These neural networks can be designed in many-to-many, many-to-one, or one-to-many formats.

In the present disclosure, the output of the artificial neural network model may be one of the elements of the action matrix. Additionally, if the artificial neural network model of the present disclosure is a neural network that processes time series input, such as a recurrent neural network, the artificial neural network model may be configured so that the output value at each time point becomes the input value at the next time point. For example, if the input of the artificial neural network model at time t is A and the output for A is B, the input of the artificial neural network model at time t+1 is B, and the input of the artificial neural network model for B at time t+1 is B. If the output is C, the input of the artificial neural network model at time t+2 may consist of C.

For time-series tasks, if an individual subaction is already scheduled, that subaction cannot be selected again. For example, in the case of a time series task that involves filling 5 holes with 5 pieces of equipment, there is a 'subaction to select the equipment' and a 'subaction to select the hole to be filled'. At this time, five actions are created by assigning a hole to be filled to each piece of equipment, and a work schedule is created. Within 5 actions, if you assign Equipment 1 to Pit 1 the first time, you cannot select Equipment 1 and Pit 1 again for the second, third, fourth, or fifth actions.

In other cases, combinations of specific sub-actions may be impossible for reasons such as physical limitations. In the example above, if the work range of equipment 4 is shorter than the other ranges and cannot fill the distant pit 2, the subaction that selects equipment 4 and the subaction that selects pit 2 cannot be combined. Therefore, the action of filling hole 2 with equipment 4 is an action that cannot be performed.

To reflect these characteristics, the action matrix constructed based on the feature set of sub-actions in the present disclosure may be an action matrix that is updated over time. Specifically, the processor 110 can update the action matrix for each time point, and among the elements present in the action matrix at each time point, elements related to actions that have already been performed and elements related to actions that cannot be performed at a specific time point are masked. You can update the action matrix by doing this.

The output of the artificial neural network model of the present disclosure is one of the elements of the action matrix, but at each time point, the artificial neural network model can generate an output value by referring to the action matrix corresponding to that time point.

Since the output of the artificial neural network model is a set of values selected from the elements of the action matrix, it is necessary to transform it into a specific action. Based on the output of the artificial neural network model, the processor 110 can output actions performed at each point in time and create a schedule for time-series work.

The processor 110 can perform reinforcement learning by executing the schedule of time-series tasks in an actual field or in a simulation environment.

Specifically, the processor 110 may calculate the cost or time required to perform an action corresponding to each point in the schedule. Thereafter, the processor 110 may generate a reward for the reinforcement learning agent corresponding to each time point based on cost or time. Additionally, the processor may generate state information of the reinforcement learning agent for each point in time. Thereafter, the processor 110 may perform learning of the reinforcement learning agent based on the action, reward, and state information performed (or planned to be performed) by the reinforcement learning agent at each time point.

The reinforcement learning agent of the present disclosure can be trained to minimize at least one of the cost or time required to perform a task.

As in the present disclosure, the entire work is performed by identifying the sub-actions that make up the individual actions of the work, creating an action set based on the feature set for the sub-action, and then planning the actions to be performed at each time through the action set. This has the effect of significantly reducing the cost or time required to do so.

Meanwhile, the artificial neural network model of the present disclosure may include a model deployed and operated on an artificial intelligence platform (eg, a Machine Learning Operations (MLOps) platform, etc.). In addition, according to various embodiments of the present disclosure, it is determined whether the artificial neural network (or artificial intelligence) model needs to be updated, and update information about the artificial neural network model is provided to an artificial intelligence platform (e.g., MLOps (Machine Learning Operations ) may be provided to users through platforms, etc.).

Specifically, the processor 110 may determine that an update to the artificial neural network model is necessary when consideration of a plurality of sub-actions to be output by the existing artificial neural network model is required. For example, when the number of welding points increases in a welding process, the processor 110 determines that subactions have been added equal to the increased number of welding points, and updates the artificial neural network model that determines the equipment to perform welding and the welding points. You may decide that you need it.

If the processor 110 determines that the artificial neural network model needs to be updated, it may provide update information for the artificial neural network model through an artificial intelligence platform (eg, MLOps platform). Specifically, the processor 110 generates an updated feature set reflecting the characteristics of additional sub-actions, generates an updated action matrix based on the updated feature set, and generates the updated action matrix through an artificial intelligence platform. can be provided.

According to an embodiment of the present disclosure, the memory 130 is a flash memory type, hard disk type, multimedia card micro type, or card type memory (e.g. (e.g. SD or -Only Memory), and may include at least one type of storage medium among magnetic memory, magnetic disk, and optical disk. The computing device 100 may operate in connection with web storage that performs a storage function of the memory 130 on the Internet. The description of the memory described above is merely an example, and the present disclosure is not limited thereto.

The network unit 150 according to an embodiment of the present disclosure includes Public Switched Telephone Network (PSTN), x Digital Subscriber Line (xDSL), Rate Adaptive DSL (RADSL), Multi Rate DSL (MDSL), and VDSL ( A variety of wired communication systems can be used, such as Very High Speed DSL), Universal Asymmetric DSL (UADSL), High Bit Rate DSL (HDSL), and Local Area Network (LAN).

In addition, the network unit 150 presented in this specification includes Code Division Multi Access (CDMA), Time Division Multi Access (TDMA), Frequency Division Multi Access (FDMA), Orthogonal Frequency Division Multi Access (OFDMA), and SC-FDMA ( A variety of wireless communication systems can be used, such as Single Carrier-FDMA) and other systems.

In the present disclosure, the network unit 150 may use any type of wired or wireless communication system.

The techniques described herein can be used in the networks mentioned above, as well as other networks.

Figure 2 is a schematic diagram showing a network function according to an embodiment of the present disclosure.

Throughout this specification, computational model, neural network, network function, and neural network may be used interchangeably. A neural network can generally consist of a set of interconnected computational units, which can be referred to as nodes. These nodes may also be referred to as neurons. A neural network consists of at least one node. Nodes (or neurons) that make up neural networks may be interconnected by one or more links.

Within a neural network, one or more nodes connected through a link may form a relative input node and output node relationship. The concepts of input node and output node are relative, and any node in an output node relationship with one node may be in an input node relationship with another node, and vice versa. As described above, input node to output node relationships can be created around links. One or more output nodes can be connected to one input node through a link, and vice versa.

In a relationship between an input node and an output node connected through one link, the value of the data of the output node may be determined based on the data input to the input node. Here, the link connecting the input node and the output node may have a weight. Weights may be variable and may be varied by the user or algorithm in order for the neural network to perform the desired function. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to the values input to the input nodes connected to the output node and the links corresponding to each input node. The output node value can be determined based on the weight.

As described above, in a neural network, one or more nodes are interconnected through one or more links to form an input node and output node relationship within the neural network. The characteristics of the neural network can be determined according to the number of nodes and links within the neural network, the correlation between the nodes and links, and the value of the weight assigned to each link. For example, if the same number of nodes and links exist and two neural networks with different weight values of the links exist, the two neural networks may be recognized as different from each other.

A neural network may consist of a set of one or more nodes. A subset of nodes that make up a neural network can form a layer. Some of the nodes constituting the neural network may form one layer based on the distances from the first input node. For example, a set of nodes with a distance n from the initial input node may constitute n layers. The distance from the initial input node can be defined by the minimum number of links that must be passed to reach the node from the initial input node. However, this definition of a layer is arbitrary for explanation purposes, and the order of a layer within a neural network may be defined in a different way than described above. For example, a layer of nodes may be defined by distance from the final output node.

The initial input node may refer to one or more nodes in the neural network through which data is directly input without going through links in relationships with other nodes. Alternatively, in a neural network network, in the relationship between nodes based on links, it may mean nodes that do not have other input nodes connected by links. Similarly, the final output node may refer to one or more nodes that do not have an output node in their relationship with other nodes among the nodes in the neural network. Additionally, hidden nodes may refer to nodes constituting a neural network other than the first input node and the last output node.

The neural network according to an embodiment of the present disclosure is a neural network in which the number of nodes in the input layer may be the same as the number of nodes in the output layer, and the number of nodes decreases and then increases again as it progresses from the input layer to the hidden layer. You can. In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer may be less than the number of nodes in the output layer, and the number of nodes decreases as it progresses from the input layer to the hidden layer. there is. In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer may be greater than the number of nodes in the output layer, and the number of nodes increases as it progresses from the input layer to the hidden layer. You can. A neural network according to another embodiment of the present disclosure may be a neural network that is a combination of the above-described neural networks.

A deep neural network (DNN) may refer to a neural network that includes multiple hidden layers in addition to the input layer and output layer. Deep neural networks allow you to identify latent structures in data. In other words, it is possible to identify the potential structure of a photo, text, video, voice, or music (e.g., what object is in the photo, what the content and emotion of the text are, what the content and emotion of the voice are, etc.) . Deep neural networks include convolutional neural networks (CNN), recurrent neural networks (RNN), auto encoders, generative adversarial networks (GAN), and restricted Boltzmann machines (RBM). machine), deep belief network (DBN), Q network, U network, Siamese network, Generative Adversarial Network (GAN), etc. The description of the deep neural network described above is only an example and the present disclosure is not limited thereto.

In one embodiment of the present disclosure, the network function may include an autoencoder. An autoencoder may be a type of artificial neural network to output output data similar to input data. The autoencoder may include at least one hidden layer, and an odd number of hidden layers may be placed between input and output layers. The number of nodes in each layer may be reduced from the number of nodes in the input layer to an intermediate layer called the bottleneck layer (encoding), and then expanded symmetrically and reduced from the bottleneck layer to the output layer (symmetrical to the input layer). Autoencoders can perform nonlinear dimensionality reduction. The number of input layers and output layers may correspond to the dimension after preprocessing of the input data. In an auto-encoder structure, the number of nodes in the hidden layer included in the encoder may have a structure that decreases as the distance from the input layer increases. If the number of nodes in the bottleneck layer (the layer with the fewest nodes located between the encoder and decoder) is too small, not enough information may be conveyed, so if it is higher than a certain number (e.g., more than half of the input layers, etc.) ) may be maintained.

A neural network may be trained in at least one of supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning. Learning of a neural network may be a process of applying knowledge for the neural network to perform a specific operation to the neural network.

Neural networks can be trained to minimize output errors. In neural network learning, learning data is repeatedly input into the neural network, the output of the neural network and the error of the target for the learning data are calculated, and the error of the neural network is transferred from the output layer of the neural network to the input layer in the direction of reducing the error. This is the process of updating the weight of each node in the neural network through backpropagation. In the case of teacher learning, learning data in which the correct answer is labeled in each learning data is used (i.e., labeled learning data), and in the case of non-teacher learning, the correct answer may not be labeled in each learning data. That is, for example, in the case of teacher learning regarding data classification, the learning data may be data in which each learning data is labeled with a category. Labeled training data is input to the neural network, and the error can be calculated by comparing the output (category) of the neural network with the label of the training data. As another example, in the case of non-teachable learning for data classification, the error can be calculated by comparing the input training data with the neural network output. The calculated error is backpropagated in the reverse direction (i.e., from the output layer to the input layer) in the neural network, and the connection weight of each node in each layer of the neural network can be updated according to backpropagation. The amount of change in the connection weight of each updated node may be determined according to the learning rate. The neural network's calculation of input data and backpropagation of errors can constitute a learning cycle (epoch). The learning rate may be applied differently depending on the number of repetitions of the learning cycle of the neural network. For example, in the early stages of neural network training, a high learning rate can be used to increase efficiency by allowing the neural network to quickly achieve a certain level of performance, and in the later stages of training, a low learning rate can be used to increase accuracy.

In the learning of neural networks, the training data can generally be a subset of real data (i.e., the data to be processed using the learned neural network), and thus the error for the training data is reduced, but the error for the real data is reduced. There may be an incremental learning cycle. Overfitting is a phenomenon in which errors in actual data increase due to excessive learning on training data. For example, a phenomenon in which a neural network that learned a cat by showing a yellow cat fails to recognize that it is a cat when it sees a non-yellow cat may be a type of overfitting. Overfitting can cause errors in machine learning algorithms to increase. To prevent such overfitting, various optimization methods can be used. To prevent overfitting, methods such as increasing the learning data, regularization, dropout to disable some of the network nodes during the learning process, and use of a batch normalization layer can be applied. You can.

Figure 3 is a conceptual diagram showing a reinforcement learning model according to an embodiment of the present disclosure.

Reinforcement learning is an artificial neural network based on the reward calculated for the action selected by the artificial neural network model, so that the artificial neural network model can determine a better action based on the input state. It is a type of learning method that trains a model. The reinforcement learning method can be understood as a “learning method through trial and error” in that rewards are given for decisions (i.e. actions). The reward given to the artificial neural network model during the reinforcement learning process may be the cumulative reward of the results of multiple actions. Reinforcement learning creates an artificial neural network model that maximizes the reward itself or return (sum of rewards) by considering various states and rewards according to actions through learning. In the present disclosure, “reinforcement learning model” refers to a subject that determines behavior and can be used interchangeably with the term “agent.” In the technical field related to reinforcement learning, the term "Environment (Env)" or "environment model" is used to refer to "a model that returns a result considering the agent's behavior." The environment model may be a model that returns output data (e.g. state information) for given input data (e.g. control information). An environmental model may be a model in which the model structure for reaching from input to output or the causal relationship between input and output data is unknown. Agents and environments can operate by exchanging data with each other.

In FIG. 3, the reinforcement learning model 310 is a subject that determines actions based on state information and rewards and can be understood as an “agent.” In the present disclosure, status information may include current status information and next status information. Current status information and next status information can be distinguished according to the time or order in which the corresponding status information was acquired, and based on the precedence relationship, the current status information ( ) and the following status information ( ) can be named respectively.

In this disclosure, environment 330 provides “current state information” on which reinforcement learning model 310 can base its action decisions. )" can be calculated. The processor 110 provides current state information including at least one state variable from the environment 330 ( ), the current state information can be input to the reinforcement learning model 310.

In this disclosure, the processor 110 inputs current state information into the reinforcement learning model 310 and then takes action based on the reinforcement learning model 310 ( ) can be calculated. The reinforcement learning model 310 uses state information obtained from the environment 330 at a random time t ( ), the probability distribution for a plurality of selectable actions can be calculated. The processor 110 takes action based on the calculated probability distribution ( ) can be calculated. For example, the processor 110 selects the largest value among the probability distributions for a plurality of actions (action ( ) can be determined.

In this disclosure, processor 110 may input a behavior calculated based on reinforcement learning model 310 into environment 330 . The processor 110 updates the next state information from the environment 330 as a result of the input of the action ( ) and compensation ( ) can be obtained. A “reward function” that serves as a standard for the environment 330 to determine a reward, or a “transition probability distribution function” that serves as a standard for determining the next state information after the environment 330 receives an action from the reinforcement learning model 310. Reinforcement learning when is known can be called “model-based” reinforcement learning. On the other hand, reinforcement learning when the reward function of the environment 330 and the transition probability distribution function of the environment 330 are unknown may be called “model-free” reinforcement learning.

The reinforcement learning model according to the present disclosure can be learned based on at least one episode. In the present disclosure, “episode” may be used as a term meaning a data sequence having a series of sequences. An episode may be a data set composed of a plurality of N-tuple data containing N elements (N is a natural number greater than 1). A plurality of N-tuple data included in an episode may have a serial order. As an example of N-tuple data, when N is '4', each 4-tuple data may include current state information, action information, reward, and next state information as elements. As another example regarding N-tuple data, when N is '5', each 5-tuple data includes current state information, action information corresponding to the current state information, reward, next state information, and next state information. Behavior information can also be included as an element.

The processor 110 according to the present disclosure performs a plurality of steps that are the same or similar to the embodiment of the above-described learning method from the initial state (t=0) to the final state (t=T). One episode can be obtained by performing it repeatedly. The final state may be derived when a preset termination condition is satisfied or when a preset number of steps are performed. The step refers to at least one action unit in which the reinforcement learning model receives a state, determines an action, and then receives a reward or updated state information for the action. The number of preset steps can be set to any natural number and, for example, can consist of 200 steps.

The processor 110 may train a reinforcement learning model based on at least one training data. For example, the processor 110 may learn a reinforcement learning model at the end of each step based on training data corresponding to each step. As another example, the processor 110 may learn a reinforcement learning model based on a training data set including training data for each of the plurality of steps at the end of each episode including a plurality of steps. As another example, after steps of a predetermined batch size are performed, the processor 110 may learn a reinforcement learning model based on a training data set including training data for each step. The batch size may be predetermined as an arbitrary natural number.

According to the present disclosure, the process of the processor 110 learning a reinforcement learning model may include modifying the weight or bias value of each node of the neural network included in the reinforcement learning model. The step of modifying the weight or bias value of each node of the neural network included in the reinforcement learning model is the same or similar to the backpropagation technique for the neural network described above with reference to FIG. 2 by the processor 110. It can be done in this way. For example, the reward ( ) is a positive number, the processor 110 performs an action (action) included in the learning data (T_t). ) can be adjusted to adjust the weight or bias value of one or more nodes included in the reinforcement learning model so that it is strengthened. At this time, one or more nodes included in the reinforcement learning model, the reinforcement learning model has state information (T_t) included in the learning data (T_t) ) and then take the above action ( ) may be a node involved in determining.

The learned reinforcement learning model 310 can determine an action in each state information so that the cumulative value (i.e. return) of rewards given from the environment 330 is maximized. The method by which the reinforcement learning model 310 determines an action may include, for example, a value-based action decision method, a policy-based action decision method, and an action decision method based on both values and policies. It can be based on at least one. The value-based action decision method is a method of determining the action that gives the highest value in each state based on a value function. Examples of value-based action decision methods may include Q-learning, Deep Q-Network (DQN), etc. The policy-based action decision method is a method of determining action based on the final return and policy function without a value function. Examples of policy-based action decision methods may include the Policy Gradient technique. The behavior decision method based on both value and policy is a method of determining the agent's behavior by learning in a way that the value function evaluates the behavior when the policy function determines the behavior. Methods for determining actions based on both values and policies may include, for example, Actor-Critic Algorithm, Soft Actor-Critic Algorithm, A2C Algorithm, A3C Algorithm, etc.

The specific descriptions regarding learning of the reinforcement learning model described above are only for explanation and do not limit the present disclosure.

Figure 4 is a flowchart showing a process for creating a schedule for a time-series task according to an embodiment of the present disclosure.

According to FIG. 4, the process of generating a schedule of a time-series task according to an embodiment of the present disclosure includes identifying a plurality of sub-actions constituting an individual action (S410) and a feature set for the plurality of sub-actions. A step of generating (S420), a step of generating an action matrix based on a feature set for each sub-action (S430), and a step of generating a schedule for performing a task based on the action matrix using an artificial neural network model. (S440) may be included.

In step S410, the processor 110 may identify a plurality of sub-actions that constitute individual actions included in the task. The definition of sub-actions and specific examples of sub-actions were described above with reference to FIG. 1.

In step S420, the processor 110 may generate a feature set for a plurality of sub-actions. Specifically, in order to generate a feature set for sub-actions, the processor 110 identifies selectable values for each sub-action and inputs the identified information into an encoder-type artificial neural network in units of sub-actions, respectively. A feature set corresponding to the sub-action of can be created.

For example, in the case of performing welding on multiple welding points included in one product using multiple welding equipment, the action at each point is the operation of which equipment is assigned to one welding point to perform welding. It may mean, and the sub-action may be ‘an operation to determine the equipment to perform welding’ and ‘an operation to determine the welding point.’

At this time, the information on the welding equipment can be expressed in the form of [Equation 1].

는 i번째 용접 장비에 대한 정보를 의미할 수 있다.At this time

may mean information about the ith welding equipment.

In the same way, information on welding points can be expressed in the form of [Equation 2].

는 j번째 용접 포인트에 대한 정보를 의미할 수 있다.At this time

may mean information about the jth welding point.

The processor 110 may generate a feature set for the welding equipment by inputting information about the welding equipment into an encoder-type artificial neural network. The feature set for welding equipment can be expressed in the form of [Equation 3].

는 i번째 용접 장비의 정보가 인코딩된 값을 의미할 수 있다.At this time

may mean a value in which information about the ith welding equipment is encoded.

In the same way, the processor 110 may generate a feature set for the weld point by inputting information about the weld point into an artificial neural network of the same artificial neural network or another encoder structure with a similar structure. The feature set for the welding point can be expressed in the form of [Equation 4].

는 j번째 용접 포인트의 정보가 인코딩된 값을 의미할 수 있다.At this time

may mean a value in which information about the jth welding point is encoded.

As in the example above, the processor 110 may input information identified for each sub-action into an encoder-type artificial neural network to generate a feature set corresponding to each sub-action.

크기의 행렬일 수 있다. 위 예시에서는 서브 액션의 수가 2이므로 액션 행렬은 2차원 행렬이었으나, 본 개시는 서브 액션의 개수를 한정하지 아니하고 다양한 유형의 작업에 대하여 액션 행렬을 생성할 수 있다.In step S430, the processor 110 may generate an action matrix based on a feature set for each sub-action. The action matrix may be the result of a matrix multiplication operation of each feature set. In an embodiment in which welding is performed on a plurality of welding points using a plurality of welding equipment, when the number of welding equipment is M and the number of welding points is N, the action matrix is a matrix of size M*1 and size of 1*N. The matrix product of the matrix of

It can be a matrix of sizes. In the above example, the number of sub-actions was 2, so the action matrix was a two-dimensional matrix. However, the present disclosure does not limit the number of sub-actions and can generate action matrices for various types of tasks.

In step S440, the processor 110 may utilize an artificial neural network model to create a schedule for performing a task based on the action matrix. When creating a schedule, the action matrix is updated at each time point, and the output value after the update can be generated by reflecting changes in the action matrix. In an embodiment in which welding is performed at a plurality of welding points using a plurality of welding equipment, the processor 110 inputs the 'START' token into an LSTM-type artificial neural network to select one element of the two-dimensional action matrix. 1 can produce output. Thereafter, the processor 110 may update the action matrix by masking elements related to the first output among the entire action matrix. Next, the processor 110 may generate a second output by inputting the first output into an LSTM type artificial neural network and selecting one element of the updated action matrix. In this way, the processor 110 can generate M output values equal to the number of welding equipment.

Since each output value is a value selected from the action matrix generated through operations between feature sets, the processor 110 can convert the value selected from the action matrix back into the form of an action. That is, the processor 110 can identify an element corresponding to each output value among the elements of the action matrix and identify the action corresponding to the corresponding element.

FIG. 5 is a conceptual diagram illustrating a process for generating an action matrix when the number of sub-actions is 2 and a process for generating a schedule based on the action matrix using an artificial neural network model according to an embodiment of the present disclosure.

Hereinafter, each configuration of FIG. 6 will be described focusing on 'an embodiment of performing welding at multiple welding points using multiple welding equipment' in which the number of sub-actions is 2.

The processor 110 may generate a feature set 712 about the welding equipment by inputting information 711 about the welding equipment into the first encoder 710. Additionally, the processor 110 may generate a feature set 722 for the welding point by inputting information 721 about the welding point into the second encoder 720.

크기의 2차원 행렬일 수 있다. 이 때, 액션 행렬(730)의 요소인

는 i번째 장비의 정보 및 j번째 용접 포인트의 정보와 관련된 값일 수 있다.Afterwards, the processor 110 may generate an action matrix 730 based on information 711 about welding equipment and information 721 about welding points. If the number of welding equipment is M and the number of welding points is N, the action matrix is

It can be a two-dimensional matrix of size. At this time, the elements of the action matrix 730 are

May be a value related to information on the ith equipment and information on the jth welding point.

를 출력한 경우, 이는 t 시점에서 액션 행렬(730)의 요소 중

를 선택하는 것을 의미할 수 있다. The processor 110 may generate an output value using an artificial neural network model at a specific point in time (or in a specific order). Specifically, the processor 110 may select and output one of each element of the action matrix by inputting a 'START' token to the artificial neural network model 740 at time t or in order t. For example, the artificial neural network model 740

When output, this is one of the elements of the action matrix 730 at time t.

It may mean choosing .

를 출력한 경우 프로세서(110)는 액션 행렬의 요소 중

가 선택되었음을 식별하고, 액션 행렬의 요소 중 i및 j와 관련된 요소를 마스킹하여 액션 행렬(730)을 업데이트할 수 있다.Thereafter, the processor 110 may mask elements related to actions that have already been performed among the elements present in the action matrix and elements related to actions that cannot be performed at the specific point in time among the elements present in the action matrix 730. For example, at time t, the artificial neural network model 740

When output, the processor 110 selects one of the elements of the action matrix.

It is possible to identify that is selected and update the action matrix 730 by masking elements related to i and j among the elements of the action matrix.

By repeating the process of generating the output value of the artificial neural network model 740 and updating the action matrix 730 as above, the output values can be obtained sequentially at each time point (or sequence), and finally a set of output values in the form of a sequence can be obtained. You can. The output value set may have the form of [Equation 5].

Finally, the processor 110 can create a schedule for all tasks by identifying actions corresponding to each output value.

In one embodiment of the present disclosure, processor 110 may generate an output value of an artificial neural network model using an attention mechanism. For example, the processor 110 identifies a hidden state for the input value of the artificial neural network at a specific point in time (or sequence) and uses an attention mechanism based on the hidden state and the action matrix 730 to identify the hidden state. You can generate output values at a specific point in time.

Meanwhile, a computer-readable medium storing a data structure is disclosed according to an embodiment of the present disclosure.

Data structure can refer to the organization, management, and storage of data to enable efficient access and modification of data. Data structure can refer to the organization of data to solve a specific problem (e.g., retrieving data, storing data, or modifying data in the shortest possible time). A data structure may be defined as a physical or logical relationship between data elements designed to support a specific data processing function. Logical relationships between data elements may include connection relationships between user-defined data elements. Physical relationships between data elements may include actual relationships between data elements that are physically stored in a computer-readable storage medium (e.g., a persistent storage device). A data structure may specifically include a set of data, relationships between data, and functions or instructions applicable to the data. Effectively designed data structures allow computing devices to perform computations while minimizing the use of the computing device's resources. Specifically, computing devices can increase the efficiency of operations, reading, insertion, deletion, comparison, exchange, and search through effectively designed data structures.

Data structures can be divided into linear data structures and non-linear data structures depending on the type of data structure. A linear data structure may be a structure in which only one piece of data is connected to another piece of data. Linear data structures may include List, Stack, Queue, and Deque. A list can refer to a set of data that has an internal order. The list may include a linked list. A linked list may be a data structure in which data is connected in such a way that each data is connected in a single line with a pointer. In a linked list, a pointer may contain connection information to the next or previous data. Depending on its form, a linked list can be expressed as a singly linked list, a doubly linked list, or a circularly linked list. A stack may be a data listing structure that allows limited access to data. A stack can be a linear data structure in which data can be processed (for example, inserted or deleted) at only one end of the data structure. Data stored in the stack may have a data structure (LIFO-Last in First Out) where the later it enters, the sooner it comes out. A queue is a data listing structure that allows limited access to data. Unlike the stack, it can be a data structure (FIFO-First in First Out) where data stored later is released later. A deck can be a data structure that can process data at both ends of the data structure.

A non-linear data structure may be a structure in which multiple pieces of data are connected behind one piece of data. Nonlinear data structures may include graph data structures. A graph data structure can be defined by vertices and edges, and an edge can include a line connecting two different vertices. Graph data structure may include a tree data structure. A tree data structure may be a data structure in which there is only one path connecting two different vertices among a plurality of vertices included in the tree. In other words, it may be a data structure that does not form a loop in the graph data structure.

Throughout this specification, computational model, neural network, network function, and neural network may be used interchangeably. Below, it is described in a unified manner as a neural network. Data structures may include neural networks. And the data structure including the neural network may be stored in a computer-readable medium. Data structures including neural networks also include data preprocessed for processing by a neural network, data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, neural network It may include a loss function for learning. A data structure containing a neural network may include any of the components disclosed above. In other words, the data structure including the neural network includes preprocessed data for processing by the neural network, data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, neural network It may be configured to include all or any combination of the loss function for learning. In addition to the configurations described above, a data structure containing a neural network may include any other information that determines the characteristics of the neural network. Additionally, the data structure may include all types of data used or generated in the computational process of a neural network and is not limited to the above. Computer-readable media may include computer-readable recording media and/or computer-readable transmission media. A neural network can generally consist of a set of interconnected computational units, which can be referred to as nodes. These nodes may also be referred to as neurons. A neural network consists of at least one node.

The data structure may include data input to the neural network. A data structure containing data input to a neural network may be stored in a computer-readable medium. Data input to the neural network may include learning data input during the neural network learning process and/or input data input to the neural network on which training has been completed. Data input to the neural network may include data that has undergone pre-processing and/or data subject to pre-processing. Preprocessing may include a data processing process to input data into a neural network. Therefore, the data structure may include data subject to preprocessing and data generated by preprocessing. The above-described data structure is only an example and the present disclosure is not limited thereto.

The data structure may include the weights of the neural network. (In this specification, weights and parameters may be used with the same meaning.) And the data structure including the weights of the neural network may be stored in a computer-readable medium. A neural network may include multiple weights. Weights may be variable and may be varied by the user or algorithm in order for the neural network to perform the desired function. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to the values input to the input nodes connected to the output node and the links corresponding to each input node. Based on the weight, the data value output from the output node can be determined. The above-described data structure is only an example and the present disclosure is not limited thereto.

As an example and not a limitation, the weights may include weights that are changed during the neural network learning process and/or weights for which neural network learning has been completed. Weights that change during the neural network learning process may include weights that change at the start of the learning cycle and/or weights that change during the learning cycle. Weights for which neural network training has been completed may include weights for which a learning cycle has been completed. Therefore, the data structure including the weights of the neural network may include weights that are changed during the neural network learning process and/or the data structure including the weights for which neural network learning has been completed. Therefore, the above-mentioned weights and/or combinations of each weight are included in the data structure including the weights of the neural network. The above-described data structure is only an example and the present disclosure is not limited thereto.

The data structure including the weights of the neural network may be stored in a computer-readable storage medium (e.g., memory, hard disk) after going through a serialization process. Serialization can be the process of converting a data structure into a form that can be stored on the same or a different computing device and later reorganized and used. Computing devices can transmit and receive data over a network by serializing data structures. Data structures containing the weights of a serialized neural network can be reconstructed on the same computing device or on a different computing device through deserialization. The data structure including the weights of the neural network is not limited to serialization. Furthermore, the data structure including the weights of the neural network is a data structure to increase computational efficiency while minimizing the use of computing device resources (e.g., in non-linear data structures, B-Tree, Trie, m-way search tree, AVL tree, Red-Black Tree) may be included. The foregoing is merely an example and the present disclosure is not limited thereto.

The data structure may include hyper-parameters of a neural network. And the data structure including the hyperparameters of the neural network can be stored in a computer-readable medium. A hyperparameter may be a variable that can be changed by the user. Hyperparameters include, for example, learning rate, cost function, number of learning cycle repetitions, weight initialization (e.g., setting the range of weight values subject to weight initialization), Hidden Unit. It may include a number (e.g., number of hidden layers, number of nodes in hidden layers). The above-described data structure is only an example and the present disclosure is not limited thereto.

6 is a brief, general schematic diagram of an example computing environment in which embodiments of the present disclosure may be implemented.

Although the disclosure has been described above as being generally capable of being implemented by a computing device, those skilled in the art will understand that the disclosure can be implemented in combination with computer-executable instructions and/or other program modules that can be executed on one or more computers and/or hardware. It will be well known that it can be implemented as a combination of software.

Typically, program modules include routines, programs, components, data structures, etc. that perform specific tasks or implement specific abstract data types. Additionally, those skilled in the art will understand that the methods of the present disclosure can be used on single-processor or multiprocessor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor-based or programmable consumer electronics, etc. It will be appreciated that other computer system configurations may be implemented, including, but not limited to, each of which may operate in connection with one or more associated devices.

The described embodiments of the disclosure can also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Computers typically include a variety of computer-readable media. Computer-readable media can be any medium that can be accessed by a computer, and such computer-readable media includes volatile and non-volatile media, transitory and non-transitory media, removable and non-transitory media. Includes removable media. By way of example, and not limitation, computer-readable media may include computer-readable storage media and computer-readable transmission media. Computer-readable storage media refers to volatile and non-volatile media, transient and non-transitory media, removable and non-removable, implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Includes media. Computer readable storage media may include RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital video disk (DVD) or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage. This includes, but is not limited to, a device, or any other medium that can be accessed by a computer and used to store desired information.

A computer-readable transmission medium typically implements computer-readable instructions, data structures, program modules, or other data on a modulated data signal, such as a carrier wave or other transport mechanism. Includes all information delivery media. The term modulated data signal refers to a signal in which one or more of the characteristics of the signal have been set or changed to encode information within the signal. By way of example, and not limitation, computer-readable transmission media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above are also intended to be included within the scope of computer-readable transmission media.

An example environment 1100 is shown that implements various aspects of the present disclosure, including a computer 1102, which includes a processing unit 1104, a system memory 1106, and a system bus 1108. do. System bus 1108 couples system components, including but not limited to system memory 1106, to processing unit 1104. Processing unit 1104 may be any of a variety of commercially available processors. Dual processors and other multiprocessor architectures may also be used as processing unit 1104.

System bus 1108 may be any of several types of bus structures that may further be interconnected to a memory bus, peripheral bus, and local bus using any of a variety of commercial bus architectures. System memory 1106 includes read only memory (ROM) 1110 and random access memory (RAM) 1112. The basic input/output system (BIOS) is stored in non-volatile memory 1110, such as ROM, EPROM, and EEPROM, and is a basic input/output system that helps transfer information between components within the computer 1102, such as during startup. Contains routines. RAM 1112 may also include high-speed RAM, such as static RAM, for caching data.

Computer 1102 may also include an internal hard disk drive (HDD) 1114 (e.g., EIDE, SATA)—the internal hard disk drive 1114 may also be configured for external use within a suitable chassis (not shown). Yes - a magnetic floppy disk drive (FDD) 1116 (e.g., for reading from or writing to a removable diskette 1118), and an optical disk drive 1120 (e.g., a CD-ROM for reading the disk 1122 or reading from or writing to other high-capacity optical media such as DVDs). Hard disk drive 1114, magnetic disk drive 1116, and optical disk drive 1120 are connected to system bus 1108 by hard disk drive interface 1124, magnetic disk drive interface 1126, and optical drive interface 1128, respectively. ) can be connected to. The interface 1124 for implementing an external drive includes at least one or both of Universal Serial Bus (USB) and IEEE 1394 interface technologies.

These drives and their associated computer-readable media provide non-volatile storage of data, data structures, computer-executable instructions, and the like. For computer 1102, drive and media correspond to storing any data in a suitable digital format. Although the description of computer-readable media above refers to removable optical media such as HDDs, removable magnetic disks, and CDs or DVDs, those of ordinary skill in the art will also recognize zip drives, magnetic cassettes, flash memory cards, and cartridges. It will be appreciated that other types of computer-readable media may also be used in the example operating environment, and that any such media may contain computer-executable instructions for performing the methods of the present disclosure. .

A number of program modules may be stored in drives and RAM 1112, including an operating system 1130, one or more application programs 1132, other program modules 1134, and program data 1136. All or portions of the operating system, applications, modules and/or data may also be cached in RAM 1112. It will be appreciated that the present disclosure may be implemented on various commercially available operating systems or combinations of operating systems.

A user may enter commands and information into computer 1102 through one or more wired/wireless input devices, such as a keyboard 1138 and a pointing device such as mouse 1140. Other input devices (not shown) may include microphones, IR remote controls, joysticks, game pads, stylus pens, touch screens, etc. These and other input devices are connected to the processing unit 1104 through an input device interface 1142, which is often connected to the system bus 1108, but may also include a parallel port, an IEEE 1394 serial port, a game port, a USB port, an IR interface, It can be connected by other interfaces, etc.

A monitor 1144 or other type of display device is also connected to system bus 1108 through an interface, such as a video adapter 1146. In addition to monitor 1144, computers typically include other peripheral output devices (not shown) such as speakers, printers, etc.

Computer 1102 may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) 1148, via wired and/or wireless communications. Remote computer(s) 1148 may be a workstation, computing device computer, router, personal computer, portable computer, microprocessor-based entertainment device, peer device, or other conventional network node, and is generally connected to computer 1102. For simplicity, only memory storage device 1150 is shown, although it includes many or all of the components described. The logical connections depicted include wired/wireless connections to a local area network (LAN) 1152 and/or a larger network, such as a wide area network (WAN) 1154. These LAN and WAN networking environments are common in offices and companies and facilitate enterprise-wide computer networks, such as intranets, all of which can be connected to a worldwide computer network, such as the Internet.

When used in a LAN networking environment, computer 1102 is connected to local network 1152 through wired and/or wireless communication network interfaces or adapters 1156. Adapter 1156 may facilitate wired or wireless communication to LAN 1152, which also includes a wireless access point installed thereon for communicating with wireless adapter 1156. When used in a WAN networking environment, the computer 1102 may include a modem 1158 or be connected to a communicating computing device on the WAN 1154 or to establish communications over the WAN 1154, such as via the Internet. Have other means. Modem 1158, which may be internal or external and a wired or wireless device, is coupled to system bus 1108 via serial port interface 1142. In a networked environment, program modules described for computer 1102, or portions thereof, may be stored in remote memory/storage device 1150. It will be appreciated that the network connections shown are exemplary and that other means of establishing a communications link between computers may be used.

Computer 1102 may be associated with any wireless device or object deployed and operating in wireless communications, such as a printer, scanner, desktop and/or portable computer, portable data assistant (PDA), communications satellite, wirelessly detectable tag. Performs actions to communicate with any device or location and telephone. This includes at least Wi-Fi and Bluetooth wireless technologies. Accordingly, communication may be a predefined structure as in a conventional network or may simply be ad hoc communication between at least two devices.

Wi-Fi (Wireless Fidelity) allows connection to the Internet, etc. without wires. Wi-Fi is a wireless technology, like cell phones, that allows these devices, such as computers, to send and receive data indoors and outdoors, anywhere within the coverage area of a base station. Wi-Fi networks use wireless technology called IEEE 802.11 (a, b, g, etc.) to provide secure, reliable, and high-speed wireless connections. Wi-Fi can be used to connect computers to each other, the Internet, and wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in the unlicensed 2.4 and 5 GHz wireless bands, for example, at data rates of 11 Mbps (802.11a) or 54 Mbps (802.11b), or in products that include both bands (dual band). .

Those skilled in the art will understand that information and signals may be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols and chips that may be referenced in the above description include voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields. It can be expressed by particles or particles, or any combination thereof.

Those skilled in the art will understand that the various illustrative logical blocks, modules, processors, means, circuits and algorithm steps described in connection with the embodiments disclosed herein may be used in electronic hardware, (for convenience) It will be understood that it may be implemented by various forms of program or design code (referred to herein as software) or a combination of both. To clearly illustrate this interoperability of hardware and software, various illustrative components, blocks, modules, circuits and steps have been described above generally with respect to their functionality. Whether this functionality is implemented as hardware or software depends on the specific application and design constraints imposed on the overall system. A person skilled in the art of this disclosure may implement the described functionality in various ways for each specific application, but such implementation decisions should not be construed as departing from the scope of this disclosure.

The various embodiments presented herein may be implemented as a method, apparatus, or article of manufacture using standard programming and/or engineering techniques. The term article of manufacture includes a computer program, carrier, or media accessible from any computer-readable storage device. For example, computer-readable storage media include magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CDs, DVDs, etc.), smart cards, and flash. Includes, but is not limited to, memory devices (e.g., EEPROM, cards, sticks, key drives, etc.). Additionally, various storage media presented herein include one or more devices and/or other machine-readable media for storing information.

It is to be understood that the specific order or hierarchy of steps in the processes presented is an example of illustrative approaches. It is to be understood that the specific order or hierarchy of steps in processes may be rearranged within the scope of the present disclosure, based on design priorities. The appended method claims present elements of the various steps in a sample order but are not meant to be limited to the particular order or hierarchy presented.

The description of the presented embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Thus, the present disclosure is not limited to the embodiments presented herein but is to be interpreted in the broadest scope consistent with the principles and novel features presented herein.

Claims (22)

A method of providing an artificial neural network model that generates a schedule of time-series tasks performed by a computing device through an artificial intelligence platform,
determining whether an update to the artificial neural network model is needed; and
Providing update information about the artificial neural network model through the artificial intelligence platform.
Including,
The artificial neural network model generates an action matrix based on a set of features for a plurality of sub-actions constituting an individual action, and generates an action matrix based on the action matrix. It is a model that generates a schedule for thermal work,
Creating a schedule for performing the task based on the action matrix is:
generating an output value of the artificial neural network model for a specific point in time by referring to an action matrix updated based on the output of the artificial neural network model;
identifying an action corresponding to the output value; and
generating a schedule for performing the task based on the action determined in response to the specific point in time;
Including,
method.
According to claim 1,
The step of determining whether an update to the artificial neural network model is necessary is:
A step of determining whether consideration of additional sub-actions other than the plurality of sub-actions is necessary.
Including,
method.
According to claim 2,
The step of providing update information about the artificial neural network model through the artificial intelligence platform,
generating an updated feature set reflecting the features of the additional sub-action;
generating an updated action matrix based on the updated feature set; and
Providing the updated action matrix through the artificial intelligence platform
Including,
method.
A method performed by a computing device to create a schedule of time-series work, comprising:
Identifying a plurality of sub-actions constituting an individual action;
generating a feature set for the plurality of sub-actions;
generating an action matrix based on a feature set for the plurality of sub-actions; and
Generating a schedule for performing the task based on the action matrix using an artificial neural network model;
Including,
The step of creating a schedule for performing the task based on the action matrix is:
Generating an output value of the artificial neural network model for a specific time point by referring to an action matrix updated based on the output of the artificial neural network model;
identifying an action corresponding to the output value; and
generating a schedule for performing the task based on the action determined in response to the specific point in time;
Including,
method.
According to claim 4,
The step of generating a feature set for the plurality of sub-actions is:
Identifying selectable values for each sub-action; and
generating a feature set corresponding to each sub-action based on inputting the identified information for each sub-action into an encoder-type artificial neural network;
Including,
method.
According to claim 4,
The action matrix is,
It is a multidimensional matrix constructed based on the feature set corresponding to each sub-action,
The dimension of the action matrix is equal to the number of the plurality of sub-actions,
method.
delete delete According to claim 4,
The artificial neural network model is,
It is configured to select and output one of the elements of the updated action matrix at each point in time,
Configured so that the output value at each time point becomes the input value at the next time point,
method.
According to claim 4,
The updated action matrix is
Among the elements present in the action matrix, elements related to already performed actions; and
Among the elements present in the action matrix, an element related to an action that cannot be performed at the specific point in time;
is the masked action matrix,
method.
According to claim 4,
generating a reward for the reinforcement learning agent for each time point based on at least one of the cost or time required to perform the action corresponding to each time point;
Generating state information of a reinforcement learning agent for each time point; and
performing learning of the reinforcement learning agent based on action, reward, and state information for each time point;
Containing more,
method.
According to claim 11,
The reinforcement learning agent is,
Learning to minimize at least one of the cost or time required to perform the above task,
method.
A computer program stored in a computer-readable storage medium that causes a computing device to perform operations that generate a schedule of a time-series task consisting of a plurality of actions, the operations comprising:
Identifying a plurality of sub-actions constituting an individual action;
An operation of generating a feature set for the plurality of sub-actions;
generating an action matrix based on a set of features for the plurality of sub-actions; and
An operation of generating a schedule for performing the task based on the action matrix using an artificial neural network model;
Including,
The operation of creating a schedule for performing the task based on the action matrix is:
An operation of generating an output value of the artificial neural network model for a specific point in time by referring to an action matrix updated based on the output of the artificial neural network model;
Identifying an action corresponding to the output value; and
An operation of creating a schedule for performing the task based on the action determined in response to the specific time point;
Including,
A computer program stored on a computer-readable storage medium.
delete delete According to claim 13,
The artificial neural network model is,
It is configured to select and output one of the elements of the updated action matrix at each point in time,
Configured so that the output value at each time point becomes the input value at the next time point,
A computer program stored on a computer-readable storage medium.
According to claim 13,
The updated action matrix is
Among the elements present in the action matrix, elements related to already performed actions; and
Among the elements present in the action matrix, an element related to an action that cannot be performed at the specific point in time;
is the masked action matrix,
A computer program stored on a computer-readable storage medium.
As a computing device,
a processor containing one or more cores;
Memory;
Including,
The processor,
Identify a plurality of sub-actions that constitute an individual action,
Generate a feature set for the plurality of sub-actions,
Generating an action matrix based on a set of features for the plurality of sub-actions,
Using an artificial neural network model, a schedule for performing tasks is created based on the action matrix,
Creating a schedule for performing the task based on the action matrix is:
generating an output value of the artificial neural network model for a specific point in time by referring to an action matrix updated based on the output of the artificial neural network model;
identifying an action corresponding to the output value; and
generating a schedule for performing the task based on the action determined in response to the specific point in time;
Including,
Computing device.
delete delete According to claim 18,
The artificial neural network model is,
It is configured to select and output one of the elements of the updated action matrix at each point in time,
Configured so that the output value at each time point becomes the input value at the next time point,
Computing device.
According to claim 18,
The updated action matrix is
Among the elements present in the action matrix, elements related to already performed actions; and
Among the elements present in the action matrix, an element related to an action that cannot be performed at the specific point in time;
is the masked action matrix,
Computing device.

Images