KR20230166002A - Method for Controlling Equipment of Service Field - Google Patents

Abstract

The present invention relates to a device control method at a service site. The device control method at a service site executed through an operation server according to the present invention involves time-series sensing through S (S ≥ 1) sensors provided at the service site. Among the S sensing data and C image data captured in time series through C (C≥1) cameras installed at the site, designated D (1≤D≤(S+C)) collected data are collected. A first step of preprocessing and configuring N (1 ≤ N ≤ D) time series data modeled to enable matching within an acceptable range with T (T ≥ 1) time series models related to the time series of the process performed in the field, Based on N time series data, t(1≤t≤T) time series models with a periodicity that can be matched with a specified periodicity related to at least one process performed in the field and n(1 with a periodicity that can be matched within a specified tolerance range) A second step of checking ≤n≤N) time series data, a third step of processing the confirmed n time series data into frequency domain-based time series data corresponding to specified frequency characteristics, and the frequency Among the n time series data processed into the domain, M target stars N'(1≤N) containing n'(1≤n'≤n) time series data for M(M≥1) targets required for analysis of the specified target. M analysis artificial intelligence modules are trained by constructing a plurality of analysis learning data for each M object containing '≤N, n'⊂N') time series data in a specified structure, and the M analysis artificial intelligence modules are A plurality of predictive learning data containing analysis data containing m analysis information analyzed through m analysis artificial intelligence modules corresponding to m (1≤m≤M) objects in a specified structure are designated. The fourth step of learning the artificial intelligence module for prediction, and after learning the M analysis artificial intelligence modules and the prediction artificial intelligence module, the frequencies are collected and preprocessed through the processes corresponding to the first to third steps. Among the n time series data processed into the domain, N' time series data for each M target, including n' time series data for each M target, are assigned to the M designated M analysis artificial intelligence modules to produce M analysis information. A fifth step of confirming the predicted prediction information by substituting the analysis data including the specified m analysis information among the confirmed M pieces of analysis information into the designated prediction artificial intelligence module, and confirming the predicted prediction information between the confirmed prediction information and the designated It contains p(p≥1) pieces of feature information including m'(1≤m'≤M) pieces of analysis information, and is designated e related to the feature information among E(E≥1) devices provided at the service site. A sixth step of configuring a plurality of control learning data containing e control information set to control (1≤e≤E) devices to learn the control artificial intelligence module for controlling the devices installed in the field. Wow, after learning of the control artificial intelligence module, when the specified target is analyzed through the process corresponding to the first to fifth steps and the predicted p feature information is confirmed, the p feature information is transferred to the control artificial intelligence module. A seventh step of checking the e control information for controlling the e devices provided at the site by substituting into the e devices, and performing a procedure for controlling the e devices provided at the site using the e control information. Includes the 8th step.

Description

{Method for Controlling Equipment of Service Field}

The present invention relates to a method of controlling devices at a service site, and models N (N≥1) time series data by preprocessing collected data collected time-series through various sensors or cameras provided at the service site, and sets out a series of periodic characteristics. n (1≤n≤N) time series data with are processed into the frequency domain, and among the n time series data processed into the frequency domain, multiple analysis learning data for each M object are configured to create M analysis artificial intelligence modules. Learn, and include analysis data containing m analysis information analyzed through m analysis artificial intelligence modules corresponding to designated m (1≤m≤M) targets among the M analysis artificial intelligence modules in a specified structure. After configuring a plurality of prediction learning data to train a designated prediction artificial intelligence module, analysis is performed by substituting the time series data for each target among the collected, preprocessed, and processed time series data into the frequency domain into the analysis artificial intelligence module. Confirm the information, input the analysis data including the analysis information into the designated artificial intelligence module for prediction, check the predicted prediction information, and provide the service site related to the confirmed prediction information and characteristic information including the designated analysis information. The carpenter's control learning data containing the control information of the device is configured to train the control artificial intelligence module for controlling the device, and the designated object is analyzed and the predicted characteristic information is applied to the control artificial intelligence module to be delivered to the site. It relates to a method of controlling a device installed at the site by checking control information for controlling the device.

In general, 1st and 2nd generation smart farms minimize the impact of various environmental changes outside the greenhouse and set user control settings for the greenhouse's facilities and equipment to provide an optimal cultivation and watering environment for crops inside the greenhouse. It is operated in this way.

However, the control of greenhouse facilities and equipment to provide an optimal cultivation and positive water environment for crops in the greenhouse is realistically difficult to systematize standardized cultivation knowledge due to the variety of greenhouse structures and equipment operation methods, and it is difficult to systematize standardized cultivation knowledge, and it is difficult to systematize the user's knowledge and the target cultivation environment. Because it largely relies on experience and ability, there may be significant differences between farms. In addition, in order to accurately respond to situations where crop damage is expected, such as pests and diseases, physiological disorders, and preparation for abnormal weather, it is necessary to rely on the judgment and decision-making of experts (consultants). You can't help but rely heavily on it.

Meanwhile, as prior art, Republic of Korea Patent Publication No. 10-2021-015853 (published on December 31, 2021) relates to a system and method for generating farmland cultivation maps for artificial intelligence-based agricultural robots, using aerial images. After performing polygon rendering, the boundaries between cultivated fields are identified, the actual cultivable space according to the type of crop is identified, a cultivation map is created, and then provided to the robot agricultural machinery, so that cultivation work is accurately performed even in the cultivable area near the border between cultivated fields. The goal is to make it happen.

However, the above prior art performs polygon rendering using aerial images, generates a cultivation map that accurately identifies the boundary between farmlands through the density difference between polygons, and then transmits it to a robotic agricultural machine to place crops at the exact location through the robotic agricultural machine. It is just cultivating, and various environments for actual cultivation management in smart farms (cultivation environment management, water management, crop growth and physiological management, pest and physiological disorder management, agricultural work management, production and quality management, energy management, etc. ) could not be processed.

Therefore, while overcoming the situation where existing farms strongly prefer face-to-face consulting methods and have relatively low participation or satisfaction with non-face-to-face and online consulting projects, non-face-to-face online consulting for domestic facility farmers is possible under the global pandemic situation. There is a need to find new ways to meet needs and demands.

The purpose of the present invention is to model N (N ≥ 1) time series data by preprocessing collected data collected time series through various sensors or cameras provided at service sites, and to model N (1 ≤ n ≤ N) time series data are processed into the frequency domain, and among the n time series data processed into the frequency domain, multiple analysis learning data for each M object are formed to learn M artificial intelligence modules for analysis, and M artificial intelligence modules for analysis are formed. Configures a plurality of predictive learning data containing analysis data containing m analysis information analyzed through m analysis artificial intelligence modules corresponding to m (1≤m≤M) specified targets among the intelligence modules in a specified structure. After learning the designated artificial intelligence module for prediction, the time series data for each designated target among the time series data collected, preprocessed, and processed into the frequency domain is substituted into the artificial intelligence module for analysis to check the analyzed analysis information. Confirms the predicted prediction information by substituting the analysis data included in the designated prediction artificial intelligence module, and includes control information of the device provided at the service site related to characteristic information including the confirmed prediction information and the designated analysis information. Control to control the device installed in the field by configuring the carpenter's control learning data to learn the control artificial intelligence module for controlling the device, analyzing the designated target, and assigning the predicted feature information to the control artificial intelligence module. It provides a device control method at a service site that verifies information and controls devices installed at the site.

The device control method at a service site executed through an operation server according to the present invention includes S sensing data sensed in time series through S (S ≥ 1) sensors provided at the service site and C (C) provided at the site. T related to the time series of the process performed at the site by collecting and preprocessing the designated D (1≤D≤(S+C)) collected data among C image data captured in time series through C≥1) cameras. A first step of configuring (T≥1) time series models and N (1≤N≤D) time series data modeled to be able to match within an allowable range, and at least performed in the field based on the N time series data. A second method that identifies t(1≤t≤T) time series models with periodicities that can match the specified periodicity related to one process and n(1≤n≤N) time series data with periodicities that can match within the specified tolerance range. A third step of processing the confirmed n time series data into frequency domain-based time series data corresponding to specified frequency characteristics, and analysis of a designated target among the n time series data processed into the frequency domain. N'(1≤N'≤N, n'⊂N') time series data for M objects including n'(1≤n'≤n) time series data for M(M≥1) objects required for M artificial intelligence modules for analysis are trained by constructing a plurality of analysis learning data for each M object containing a specified structure, and among the M artificial intelligence modules for analysis, the designated m (1≤m≤M) objects are used. A fourth step of training a designated artificial intelligence module for prediction by constructing a plurality of prediction learning data containing analysis data including m analysis information analyzed through corresponding m analysis artificial intelligence modules in a designated structure; After learning the M artificial intelligence modules for analysis and artificial intelligence modules for prediction, designated M targets among the n time series data collected and preprocessed through the processes corresponding to the first to third steps and processed into the frequency domain. Confirm M analysis information analyzed by substituting M objects containing time series data for n' stars into the designated M analysis artificial intelligence modules, and specify m among the confirmed M analysis information. A fifth step of confirming the predicted prediction information by substituting analysis data containing analysis information into a designated prediction artificial intelligence module, and confirming the predicted prediction information and the designated m' (1≤m'≤M) analysis information. Contains p (p≥1) pieces of characteristic information, and controls designated e (1≤e≤E) devices related to the feature information among E (E≥1) devices provided at the service site. A sixth step of configuring a plurality of control learning data containing e sets of control information to learn a control artificial intelligence module for controlling the device installed in the field, and after learning the control artificial intelligence module, When the designated object is analyzed and the predicted p feature information is confirmed through the process corresponding to steps 1 to 5, the p feature information is substituted into the control artificial intelligence module to control the e devices installed at the site. It may include a seventh step of checking e pieces of control information for

According to the present invention, the preprocessing may include an averaging procedure to reveal the representativeness of a periodic pattern or a certain time interval of the D collected data with a specified periodicity.

According to the present invention, the preprocessing may include a procedure for correcting temporal influence corresponding to bias weighted by data from a previous time on data corresponding to a designated independent variable among the D collected data.

According to the present invention, the T time series models may include a time series model that can match the time series characteristics of a process performed through a device provided at a service site.

According to the present invention, the N pieces of time-series data may include time-domain-based time-series data based on time-series characteristics of collected data generated by time-series sensing or time-series photography.

According to the present invention, the periodicity is a periodicity corresponding to at least one process repeatedly performed at regular intervals through a device provided at the service site, and a schedule related to at least one process to be performed through a device provided at the service site. periodicity, which corresponds to a change in the daily interval (or the interval between the Earth's rotations), a periodicity which corresponds to a change in a trend over a specified period, a periodicity which corresponds to a seasonal change, a periodicity which corresponds to a change in the yearly interval (or the interval between the Earth's revolutions). It may include at least one periodicity among the periodicities.

According to the present invention, the t time series models include a time series that matches the time series characteristics of the process performed in the field and a periodicity that includes multiple periodicities that match the periodic characteristics of the process performed, including a designated period. Can include time series models.

According to the present invention, the second step analyzes the N time series data based on at least one periodicity time series model to generate n time series data with a periodicity that can match at least one specified periodicity time series model and a specified tolerance range. It may include a verification step.

According to the present invention, the second step is to apply a specified exponential smoothing method to the time series data in the case of time series data in which the data change according to time is slower than a specified reference value, and then at least one periodic time series model and within a specified tolerance range are performed. It may include the step of identifying n time series data with matchable periodicity.

According to the present invention, the second step analyzes the N time series data based on at least one periodic time series model, and corresponds to a random element or residual/remainer element included in the time series data. It may include the step of separately analyzing the data to identify at least one periodicity time series model and n time series data with a periodicity that can be matched within a specified allowable range.

According to the present invention, the frequency feature may include a model-based period (or frequency) feature corresponding to a time series model with periodicity.

According to the present invention, the frequency characteristic may include a data-based period (or frequency) confirmed by reading time series data with periodicity.

According to the present invention, the frequency feature is statistical processing of the feature of the model-based cycle (or frequency) corresponding to a time series model with periodicity and the feature of the data-based cycle (or frequency) confirmed by reading time series data with periodicity. Alternatively, it may include characteristics of the period (or frequency) confirmed through numerical analysis processing.

According to the present invention, the n time series data processed into the frequency domain are processed into a frequency domain corresponding to a periodicity that matches the periodic characteristics of the process performed including the period designated at the site and a frequency characteristic that can be matched within an allowable range. can include time series data.

According to the present invention, the n' time series data are processed into a frequency domain corresponding to a periodicity that matches the periodic characteristics of the process performed, including the period specified in the field, and a frequency characteristic that can be matched within an allowable range. Among the data, at least one time series data required for analysis of a specified target may be included.

According to the present invention, the N' time series data includes n' time series data required for analysis of a designated target among the n time series data processed into the frequency domain, and analysis of a designated target among the (N-n) time series data. It may include i(1≤i≤(N-n)) time series data required for .

According to the present invention, the fourth step may further include preprocessing the N' time series data so that they can be used as learning data for designated M artificial intelligence modules.

According to the present invention, the preprocessing includes a data change procedure for changing the character or category attribute value of at least one time series data included in the N' time series data into numeric data, and the N' time series data included in the N' time series data. It may include at least one data scaling procedure that adjusts the range difference between each time series data to within a specified range.

According to the present invention, the preprocessing equalizes the periodicity of at least one time series data included in the N' time series data to a certain time period in a specified time range, thereby producing the result of the next data through k (k≥1) previous data. It may include procedures for processing values so that they can be derived.

According to the present invention, the analysis learning data includes the n' time series data as data corresponding to observation features or feature vectors for artificial intelligence learning, and observation results related to the n' time series data. Alternatively, it may include j(i≥1) pieces of data corresponding to the label.

According to the present invention, the analysis learning data includes the n' time series data and includes (1 ≤ i ≤ (N-n)) time series data required for analysis of a designated target among the (N-n) time series data. Includes n'+i) time series data as data corresponding to observation features or feature vectors for artificial intelligence learning, and observation results related to the designated time series data among the (n'+i) time series data, or It may contain j(i≥1) pieces of data corresponding to the label.

According to the present invention, the analysis data includes m analysis information analyzed through the m analysis artificial intelligence modules, and among the n time series data, n" (1≤n"≤n) required for the prediction It can contain time series data.

According to the present invention, the predicted learning data includes the analysis data as data corresponding to observation features or feature vectors for artificial intelligence learning, and the observation result or label predicted based on the analysis data ( It may contain j(i≥1) pieces of data corresponding to Label).

According to the present invention, the fifth step may further include preprocessing the N' time series data for each of the M objects so that they can be substituted into the learned M artificial intelligence modules.

According to the present invention, the feature information includes feature information corresponding to the results of artificial intelligence-based analysis and prediction using frequency domain-based time series data, artificial intelligence-based analysis using multiple frequency domain-based time series data and time domain-based time series data, and It may include feature information corresponding to the prediction result.

According to the present invention, the sixth step may further include performing a procedure for controlling the e devices installed at the site using the e control information.

According to the present invention, in the sixth step, when the e devices installed at the site are controlled using the e control information, p pieces of characteristic information and e pieces of characteristic information related to the control of the e devices provided at the site are used. A step of checking actual measurement information for m objects measured currently or after a certain period of time with respect to a combination of control information, and comparing and analyzing the p feature information and the confirmed actual measurement information for m objects to determine the e devices. A step of generating feedback information corresponding to control information to improve control-related efficiency, and configuring control learning data including the p pieces of feature information and the generated feedback information to further train the artificial intelligence module for control. It may further include.

According to the present invention, the actual measurement information is actual measurement of at least one object among production volume, production quality, occurrence of abnormality, occurrence of failure, and occurrence of environmental damage corresponding to the result of controlling e devices using the e control information. May contain information.

According to the present invention, the feedback information is designated e' (1≤e'≤e) pieces of control information for controlling the e devices provided in the field in order to improve efficiency related to the control of the e devices. It may include 'e' pieces of control information in which the control information is adjusted ex post.

According to the present invention, the control learning data includes the p pieces of feature information as data corresponding to observation features or feature vectors for artificial intelligence learning, and an observation result including the e pieces of control information or It may contain j(i≥1) pieces of data corresponding to the label.

According to the present invention, the eighth step is to control the e devices provided at the site using the e control information confirmed through the artificial intelligence module for control, and control the e devices provided at the site. A step of confirming actual measurement information for m objects measured currently or after a certain period of time with respect to a combination of related p feature information and e control information, and combining the p feature information and the confirmed actual measurement information for m objects. Comparatively analyzing and generating feedback information corresponding to control information for improving efficiency related to control of the e devices, and configuring control learning data including the p feature information and the generated feedback information for the control. A step of additionally learning the artificial intelligence module may be further included.

According to the present invention, there is an advantage of being able to predict future data such as facility preventive maintenance, marketing evidence, and environmental change prediction by using previous values in the form of past, present, and future through a device control method at a service site.

In addition, according to the present invention, there is an advantage of being able to predict abnormal cause values, such as predictive maintenance of equipment, prediction of crop growth/physiology, prediction of causes of pests and diseases, and maintenance of IoT environment by detecting patterns with repeated periodic characteristics.

1 is a diagram showing the configuration of a device control system at a service site according to an implementation method of the present invention.
Figure 2 is a flow chart showing a process of processing time series data confirmed to have a periodicity that can match the specified periodicity related to the process into frequency domain-based time series data according to the implementation method of the present invention.
Figure 3 is a flowchart showing the process of checking analysis information and prediction information using an artificial intelligence module learned according to the implementation method of the present invention.
Figure 4 shows control information for controlling the device installed in the field by learning the control artificial intelligence module for controlling the device installed in the field according to the implementation method of the present invention and then substituting characteristic information into the control artificial intelligence module. This is a flowchart showing the process of checking and controlling devices in the field.

Hereinafter, the operating principle of a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings and description. However, the drawings shown below and the description below are for preferred implementation methods among various methods for effectively explaining the characteristics of the present invention, and the present invention is not limited to the drawings and description below.

That is, the following examples correspond to preferred union examples among the numerous examples of the present invention, and in the following examples, specific configurations (or steps) are omitted, or specific configurations (or steps) are omitted. An embodiment of dividing an implemented function into a specific configuration (or step), or an embodiment of integrating a function implemented in two or more configurations (or steps) into one configuration (or step), of a specific configuration (or step) It is clearly stated that all embodiments that replace the operation sequence fall within the scope of the present invention, even if not specifically mentioned in the following embodiments. Therefore, based on the examples below, it is clearly stated that various embodiments corresponding to subsets or complements can be divided retroactively to the filing date of the present invention.

Additionally, in the following description of the present invention, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the gist of the present invention, the detailed description will be omitted. The terms described below are terms defined in consideration of functions in the present invention, and may vary depending on the intention or custom of the user or operator. Therefore, the definition should be made based on the overall content of the present invention.

As a result, the technical idea of the present invention is determined by the claims, and the following examples are a means to efficiently explain the advanced technical idea of the present invention to those skilled in the art to which the present invention pertains. It's just that.

Figure 1 is a diagram showing the configuration of a device control system at a service site according to an implementation method of the present invention.

In more detail, Figure 1 models N (N ≥ 1) time series data by pre-processing collected data collected in time series through various sensors or cameras installed at service sites, and n (1 ≤ 1) with a series of periodic characteristics. n≤N) time series data are processed into the frequency domain, and among the n time series data processed into the frequency domain, multiple analysis learning data for each M object are formed to train M analysis artificial intelligence modules and M analysis. A plurality of predictive learning data containing analysis data containing m analysis information analyzed through m analysis artificial intelligence modules corresponding to m (1≤m≤M) objects specified among the artificial intelligence modules for use in a specified structure. After configuring and learning the designated artificial intelligence module for prediction, the time series data for each designated target among the collected, preprocessed, and processed time series data in the frequency domain is substituted into the artificial intelligence module for analysis to check and analyze the analyzed analysis information. By substituting the analysis data containing the information into the designated artificial intelligence module for prediction and confirming the predicted prediction information, the control information of the device provided at the service site related to the characteristic information including the confirmed prediction information and the designated analysis information is collected. Construct the carpenter's control learning data to learn the control artificial intelligence module for controlling the device, analyze the designated target, and substitute the predicted characteristic information into the control artificial intelligence module to control the device provided at the site. This shows the configuration of a system that checks control information for and controls devices installed in the field. Those skilled in the art will refer to Figure 1 and/or modify the system Although various implementation methods can be inferred (e.g., implementation methods in which some components are omitted, subdivided, or combined), the present invention includes all the implementation methods inferred above, and is shown in Figure 1. Its technical features are not limited solely to the implemented method.

The system of the present invention includes an operation server 100 that controls devices installed at service sites using time series analysis/prediction, and a user terminal 150 that receives the result information from the operation server 100. . Meanwhile, the operating server 100 may be implemented in the form of at least one or a combination of two or more of the following: an independent server, a combination of two or more servers, and a server program installed and run on an equipped server. The present invention is not limited by the method of implementation.

The user terminal 150 is a general term for computer devices used by users that are related to the site and check the analysis information and prediction information provided through the operation server 100, and are wired terminals such as computers and laptops connected to a wired network. It may include at least one of a wireless terminal such as a mobile phone, smartphone, tablet PC, or laptop connected to a wireless network.

Preferably, the user terminal 150 has an application or program installed and running that performs at least one procedure specified for the function of receiving result information for a specified analysis or prediction from the operation server 100. desirable.

Meanwhile, referring to Figure 1, the operation server 100 includes a data configuration unit 105 that configures time series data collected and modeled through sensors or cameras provided at the service site, and at least one function performed at the site. A data confirmation unit 110 that checks time series data with a periodicity that can match the specified periodicity related to the process, a data processing unit 115 that processes the time series data into frequency domain-based time series data, and a plurality of data for each designated target. An artificial intelligence learning unit that configures analysis learning data to learn an artificial intelligence module for analysis, and configures a plurality of prediction learning data containing analysis information analyzed through the analysis artificial intelligence module to learn a designated prediction artificial intelligence module. (120) and, among the time series data collected, preprocessed, and then processed into the frequency domain, the time series data for each specified target is substituted into the artificial intelligence module for analysis to check the analyzed analysis information, and the analysis data including the analysis information is used to make the specified prediction. An information confirmation unit 125 that confirms prediction information predicted by inputting it into an artificial intelligence module, and control information of devices installed at the service site related to feature information including the confirmed prediction information and designated analysis information. A control module learning unit 130 that configures control learning data of a carpenter to learn a control artificial intelligence module for controlling the device, and analyzes a designated object and substitutes the predicted feature information into the control artificial intelligence module to control the device at the site. It may be configured to include a control information confirmation unit 135 that checks control information for controlling a device provided in the field, and a device control unit 140 that controls the device provided in the field using the control information.

According to Figure 1, the data configuration unit 105 includes S sensing data sensed in time series through S (S ≥ 1) sensors provided at the service site and C (C ≥ 1) provided at the service site. Among C image data captured in time series through cameras, designated D (1≤D≤(S+C)) collected data are collected and preprocessed to determine T(T≥1) related to the time series of the process performed at the site. ) time series models and N (1≤N≤D) time series data modeled to enable matching within the allowable range can be configured.

Typically, the preprocessing of the collected data involves any or a combination of two or more of data cleaning, data transformation, data filtering, data integration, and data reduction at least once. It may include performing

Here, the data purification includes a procedure for filling missing data with default values or replacing them with average/median values, a procedure for identifying and correcting or removing abnormal data corresponding to an abnormal state, and a procedure for smoothing noisy data. It may include any one or a combination of two or more.

In addition, the data conversion may include a procedure for converting data into a form that is easy to analyze through any one or a combination of two or more of normalization, aggregation, summarization, and hierarchy creation. You can.

Additionally, the data filtering may include any one or a combination of two or more of an error data discovery procedure, an error data correction procedure, an error data deletion procedure, a duplicate data confirmation procedure, a duplicate data correction procedure, and a duplicate data deletion procedure.

Additionally, the data integration may include a procedure for integrating similar data or linked data to facilitate data analysis.

Additionally, the data reduction may include a procedure for removing unused data from collected data.

Meanwhile, according to the implementation method of the present invention, the preprocessing of the collected data may include an averaging procedure to reveal the representativeness of a periodic pattern or a certain time interval of the collected data with a specified periodicity among the D collected data.

In addition, according to the implementation method of the present invention, the preprocessing of the collected data is a procedure of correcting the temporal influence corresponding to the bias weighted by data of the previous time with respect to the data corresponding to the designated independent variable among the D collected data. may include.

Meanwhile, T (T ≥ 1) time series models related to the time series characteristics of the process performed at the field may include a time series model that can match the time series characteristics of the process performed through a device provided at the service site.

In addition, the N (1≤N≤D) time series data modeled to enable matching within the above allowable range is a time domain based on the time series characteristics of the collected data generated by time series sensing or time series shooting. Can include time series data based.

According to Figure 1, when N time series data is configured through the data configuration unit 105, the data confirmation unit 110 determines the information related to at least one process performed at the site based on the N time series data. You can check t(1≤t≤T) time series models with periodicity that can match the specified periodicity and n(1≤n≤N) time series data with periodicity that can match within the specified allowable range.

According to the implementation method of the present invention, the periodicity is a periodicity corresponding to at least one process repeatedly performed at regular intervals through a device provided at the service site, and at least one process to be performed through a device provided at the service site. periodicity corresponding to a schedule related to a periodicity, a periodicity corresponding to a change in a daily interval (or the Earth's rotation interval), a periodicity corresponding to a change in a trend in a specified period, a periodicity corresponding to a seasonal change, and a periodicity corresponding to a yearly interval (or the Earth's rotation interval). It may include at least one periodicity among the periodicities corresponding to change.

Here, the periodicity may not include periodicity related to simple vibration of the device sensed through a conventional vibration sensor and periodicity related to the frequency of electromagnetic waves sensed through an electromagnetic wave sensor.

According to the implementation method of the present invention, t (1 ≤ t ≤ T) time series models with periodicities that can match the specified periodicity related to the at least one process are clocks matched with the time series characteristics of the process performed in the field. It can include a periodic time series model that includes multiple periodicities that match the periodic characteristics of the process being performed, including recess and a specified period.

In addition, the n (1 ≤ n ≤ N) time series data with the periodicity may include time series data with a specified periodicity related to at least one process performed in the field and a periodicity matched within a specified tolerance range.

Additionally, the n time series data may include a periodicity time series model that can match a specified periodicity related to at least one process performed in the field and time series data with a periodicity that matches within a specified tolerance range.

Meanwhile, according to the implementation method of the present invention, the data confirmation unit 110 analyzes the N time series data based on at least one periodic time series model and can match at least one specified periodic time series model within a specified allowable range. You can check n time series data with periodicity.

In addition, according to the implementation method of the present invention, in the case of time series data where the data change due to time change is slower than a specified reference value, the data confirmation unit 110 applies a specified exponential smoothing method to the time series data and then performs at least one You can check n time series data with periodicity that can match the periodic time series model and within the specified tolerance range.

In addition, according to the implementation method of the present invention, the data confirmation unit 110 analyzes the N time series data based on at least one periodic time series model, and random elements or residuals included in the time series data (Residual/Remainder) By separating and analyzing the data corresponding to the element, it is possible to identify at least one periodicity time series model and n time series data with a periodicity that can be matched within the specified allowable range.

According to Figure 1, when n (1≤n≤N) time series data with the periodicity are confirmed through the data confirmation unit 110, the data processing unit 115 selects the confirmed n time series data. It can be processed into frequency domain-based time series data corresponding to specified frequency characteristics.

According to the implementation method of the present invention, the frequency feature may include a model-based period (or frequency) feature corresponding to a time series model with periodicity.

Additionally, according to the implementation method of the present invention, the frequency characteristic may include a data-based period (or frequency) confirmed by reading time series data with periodicity.

In addition, according to the implementation method of the present invention, the frequency feature is the characteristic of a model-based period (or frequency) corresponding to a time series model with periodicity and a data-based period (or frequency) confirmed by reading time series data with periodicity. The characteristics of may be statistically processed (e.g., average, correction, etc.) or numerical analysis process (e.g., average, correction, interpolation, etc.) may be performed to include the identified period (or frequency) characteristics.

Meanwhile, according to the implementation method of the present invention, the n time series data processed into the frequency domain correspond to a periodicity that matches the periodic characteristics of the process performed including the period specified in the field and a frequency characteristic that can be matched within an allowable range. It may include time series data processed into the frequency domain.

Additionally, the n pieces of time series data processed into the frequency domain do not include time series data related to simple vibration of the device sensed through a vibration sensor or time series data related to the frequency of electromagnetic waves sensed through an electromagnetic wave sensor.

Here, time series data related to the simple vibration of the device sensed through the vibration sensor or time series data related to the frequency of electromagnetic waves sensed through the electromagnetic wave sensor matches the periodic characteristics of the process performed, including the period designated at the site. It may not be included in the n time series data processed into the frequency domain corresponding to the periodicity and frequency characteristics that match within the allowable range.

In addition, time series data related to the simple vibration of the device sensed through the vibration sensor or time series data related to the frequency of electromagnetic waves sensed through the electromagnetic wave sensor matches the periodic characteristics of the process performed, including the period designated at the site. It can be identified or managed as time series data corresponding to a separate frequency domain that is distinguished from n time series data processed into a frequency domain corresponding to frequency characteristics that match periodicity and within an acceptable range.

According to Figure 1, when the time series data is processed into frequency domain-based time series data through the data processing unit 115, the artificial intelligence learning unit 120 selects a designated target among n time series data processed into the frequency domain. N'(1≤N'≤N, n'⊂N') for M objects containing n'(1≤n'≤n) time series data for M(M≥1) objects required for analysis. M artificial intelligence modules for analysis are trained by constructing multiple analysis learning data for each M object containing time series data in a specified structure, and among the M artificial intelligence modules for analysis, designated m (1≤m≤M) A designated prediction artificial intelligence module can be trained by constructing a plurality of prediction learning data containing analysis data containing m analysis information analyzed through m analysis artificial intelligence modules corresponding to the target in a specified structure.

According to the implementation method of the present invention, the n' time series data is processed into a frequency domain corresponding to a periodicity that matches the periodic characteristics of the process performed, including the period specified in the field, and a frequency characteristic that can be matched within an allowable range. Among n time series data, at least one time series data required for a specified analysis may be included.

Meanwhile, the N' time series data may include n' time series data required for a specified analysis among the n time series data processed into the frequency domain.

In addition, according to the implementation method of the present invention, the N' time series data includes n' time series data required for a specified analysis among the n time series data processed into the frequency domain, and a specified number of (N-n) time series data. It may contain i(1≤i≤(N-n)) time series data required for analysis.

Here, the (N-n) time series data may include time domain-based time series data.

Meanwhile, according to the implementation method of the present invention, the artificial intelligence learning unit 120 can preprocess the N' time series data so that they can be used as learning data for the designated M artificial intelligence modules.

The preprocessing may perform any one or a combination of two or more of data purification, data conversion, data filtering, data integration, and data reduction on at least one time series data included in the N' time series data at least once.

According to the implementation method of the present invention, the preprocessing includes a data change procedure for changing the character or category attribute value of at least one time series data included in the N' time series data into numeric data, the N' time series data It may include at least one data scaling adjustment procedure for adjusting the range difference between each time series data included in to within a specified range.

Here, the preprocessing equalizes the periodicity of at least one time series data included in the N' time series data to a certain time period in a specified time range to derive the result value of the next data through k (k≥1) previous data. Possible processing procedures may be included.

Meanwhile, according to the implementation method of the present invention, the analysis learning data includes the n' time series data as data corresponding to observation features or feature vectors for artificial intelligence learning, and the n' time series It may contain j(i≥1) pieces of data corresponding to observation results or labels related to the data.

In addition, according to the implementation method of the present invention, the analysis learning data includes the n' time series data, and among the (N-n) time series data, i (1 ≤ i ≤ (N-n)) time series data required for the specified analysis. Includes (n'+i) time series data as data corresponding to observation features or feature vectors for artificial intelligence learning, and includes (n'+i) time series data related to the specified time series data among the (n'+i) time series data. It may contain j(i≥1) pieces of data corresponding to observation results or labels. Here, the i time series data may include time domain-based time series data.

Additionally, the i time series data may include time domain-based time series data.

Meanwhile, according to the implementation method of the present invention, the M artificial intelligence modules may include an LSTM (Long Short-Term Memory) series algorithm optimized for time series analysis.

Meanwhile, according to the implementation method of the present invention, the analysis data may include m pieces of analysis information analyzed through the m pieces of artificial intelligence modules for analysis.

In addition, the analysis data includes m analysis information analyzed through the m analysis artificial intelligence modules, and among the n time series data, n" (1≤n"≤n) time series data required for the prediction. may include.

In addition, the predicted learning data includes the analysis data as data corresponding to observation features or feature vectors for artificial intelligence learning, and includes the observation results or labels predicted based on the analysis data. It may contain j(i≥1) corresponding data.

According to Figure 1, the information confirmation unit 125 learns the M analysis artificial intelligence modules and prediction artificial intelligence modules through the artificial intelligence learning unit 120, and then collects them through the process of Figure 2. Among n time series data that has been preprocessed and processed into the frequency domain, N' time series data for each M target, including n' time series data for each M target, are assigned to the designated M artificial intelligence modules for analysis. M pieces of analysis information can be confirmed, and analysis data including the designated m pieces of analysis information among the confirmed M pieces of analysis information can be substituted into a designated prediction artificial intelligence module to check the predicted prediction information.

Here, the n' time series data for the specified analysis are n' pieces of time series data processed into the frequency domain corresponding to the periodicity matching the periodic characteristics of the process performed including the specified period in the field and the frequency characteristics that can be matched within the allowable range. Among the time series data, at least one time series data required for the specified analysis may be included.

Additionally, the N' time series data may include n' time series data required for a specified analysis among the n time series data processed into the frequency domain.

In addition, the N' time series data includes n' time series data required for a specified analysis among the n time series data processed into the frequency domain, and i (1≤i) required for a specified analysis among the (N-n) time series data. It may contain ≤(N-n)) time series data. Here, the (N-n) time series data may include time domain-based time series data.

Here, the (N-n) time series data may include time domain-based time series data.

Meanwhile, according to the implementation method of the present invention, the information confirmation unit 125 can preprocess the N' time series data for each of the M objects so that they can be substituted into the learned M artificial intelligence modules.

Here, the preprocessing may perform any one or a combination of two or more of data purification, data conversion, data filtering, data integration, and data reduction on at least one time series data included in the N' time series data at least once. there is.

In addition, the preprocessing includes a data change procedure for changing character or category attribute values of at least one time series data included in the N' time series data into numeric data, and each time series data included in the N' time series data. It may include at least one data scaling adjustment procedure for adjusting the range difference between the data within a specified range.

In addition, the preprocessing equalizes the periodicity of at least one time series data included in the N' time series data to a certain time period in a specified time range to obtain the result value of the next data through k (k≥1) previous data. It may include procedures for processing so that it can be derived.

In addition, the information confirmation unit 125 can confirm M analysis information by substituting the N' time series data for each of the M preprocessed objects into M artificial intelligence modules.

Additionally, the analysis data may include m pieces of analysis information analyzed through the m pieces of artificial intelligence modules for analysis.

In addition, the analysis data includes m analysis information analyzed through the m analysis artificial intelligence modules, and among the n time series data, n" (1≤n"≤n) time series data required for the prediction. may include.

Meanwhile, the feature information includes result information corresponding to artificial intelligence-based analysis or prediction using frequency domain-based time series data, and results corresponding to artificial intelligence-based analysis or prediction using multiple frequency domain-based time series data and time domain-based time series data. May contain information.

Here, the feature information may not include result information corresponding to artificial intelligence-based analysis or prediction using only time domain-based time series data.

Meanwhile, according to the implementation method of the present invention, the information confirmation unit 125 stores the confirmed M analysis information and prediction information in a designated storage medium, and confirms the confirmed M pieces of analysis information and prediction information with a designated user terminal 140 related to the site. M pieces of analysis and forecast information can be provided.

According to Figure 1, the control module learning unit 130 confirms the predicted prediction information by substituting analysis data including m analysis information specified through the information confirmation unit 125 into the designated prediction artificial intelligence module. Then, it contains p(p≥1) pieces of feature information including the confirmed prediction information and specified m'(1≤m'≤M) pieces of analysis information, and E(E≥1) pieces of feature information provided at the service site. Constructs a plurality of control learning data containing e control information set to control the designated e (1≤e≤E) devices related to the characteristic information among the devices, for control of the devices installed in the field. Artificial intelligence modules can be trained.

According to the implementation method of the present invention, the control module learning unit 130 provides p feature information including the confirmed prediction information and m' analysis information to a designated expert terminal (not separately shown) to expert It is possible to check the e control information set to control the designated e devices related to the feature information among the E devices provided at the service site.

Here, the expert may include at least one of a person with a degree in a field related to the service field, a person who has obtained a designated certification in a field related to the service field, and a person with more than a specified period of experience in a field related to the service field. .

According to the implementation method of the present invention, the control module learning unit 130 can perform a procedure for controlling the e devices installed in the field using the e control information.

In addition, according to the implementation method of the present invention, when the control module learning unit 130 controls the e devices installed at the site using the e control information, the control module learning unit 130 controls the e devices installed at the site. Check the actual measurement information for m objects measured currently or after a certain period of time with respect to the combination of p feature information and e control information, and compare the p feature information with the confirmed actual measurement information for m objects. Analyze and generate feedback information corresponding to control information to improve efficiency related to control of the e devices, and configure control learning data including the p feature information and the generated feedback information to control the artificial intelligence module. can be further studied.

Here, the control module learning unit 130 may receive the actual measurement information for each of the m objects from the user terminal corresponding to the site.

In addition, the control module learning unit 130 generates/manages data corresponding to the actual measurement information for each of the m objects or generates/manages data to be used to derive the actual measurement information for each of the m objects. Data can be received, and actual measurement information for each of the m objects can be confirmed based on the received data.

In addition, the control module learning unit 130 senses data corresponding to the actual measurement information for each of the m objects among sensors provided in the field or senses data to be used to derive the actual measurement information for each of the m objects. Sensing data sensed through a sensor can be received, and actual measurement information for each m object can be confirmed based on the received sensing data.

In addition, the actual measurement information includes actual measurement information for at least one of the following: production volume, production quality, abnormality, failure, and environmental damage corresponding to the results of controlling e devices using the e control information. can do.

In addition, the feedback information includes specified e' (1≤e'≤e) pieces of control information among the e pieces of control information that controlled the e devices provided in the field in order to improve efficiency related to the control of the e devices. It may include 'e' pieces of control information adjusted after the fact.

In addition, the control learning data includes the p pieces of feature information as data corresponding to observation features or feature vectors for artificial intelligence learning, and is included in the observation result or label including the feedback information. It may contain j(i≥1) corresponding data.

In addition, the control learning data includes the p pieces of feature information as data corresponding to observation features or feature vectors for artificial intelligence learning, and produces an observation result or label containing the e pieces of control information. ) may include j(i≥1) pieces of data corresponding to ).

According to Figure 1, the control information confirmation unit 135 learns the control artificial intelligence module for controlling the device installed in the field through the control module learning unit 130, and then learns the control artificial intelligence module for controlling the device installed in the field through the control module learning unit 130. When the specified target is analyzed and the predicted p feature information is confirmed through the corresponding process, the p feature information is substituted into the control artificial intelligence module to control the e devices provided at the site. You can check the information.

According to Figure 1, when the e control information for controlling the e devices installed in the field is confirmed through the control information confirmation unit 135, the device control unit 140 uses the e control information. Thus, a procedure for controlling the e devices provided at the site can be performed.

According to the implementation method of the present invention, when the device control unit 140 controls the e devices installed at the site using the e control information confirmed through the control artificial intelligence module, the device control unit 140 controls the e devices installed at the site. Verify the actual measurement information for m objects measured currently or after a certain period of time for the combination of m feature information and e control information related to e device control, and check the p feature information and the confirmed m objects for each object. By comparing and analyzing the actual measurement information, feedback information corresponding to the control information for improving efficiency related to the control of the e devices is generated, and control learning data including the p feature information and the generated feedback information is constructed. The artificial intelligence module for control can be additionally trained.

Here, the device control unit 140 may receive the actual measurement information for each of the m objects from the user terminal corresponding to the site.

In addition, the device control unit 140 generates/manages data corresponding to the actual measurement information for each of the m objects or generates/manages data to be used to derive the actual measurement information for each of the m objects. It is possible to receive and check the actual measurement information for each of the m objects based on the received data.

In addition, the device control unit 140 includes a sensor that senses data corresponding to the actual measurement information for each of the m objects or senses data to be used to derive the actual measurement information for each of the m objects among the sensors provided in the field. Sensing data sensed through the sensor can be received, and actual measurement information for each m object can be confirmed based on the received sensing data.

Here, the actual measurement information includes actual measurement information for at least one of the following: production volume, production quality, abnormality, failure, and environmental damage corresponding to the results of controlling e devices using the e control information. can do.

In addition, the feedback information includes specified e' (1≤e'≤e) pieces of control information among the e pieces of control information that controlled the e devices provided in the field in order to improve efficiency related to the control of the e devices. It may include 'e' pieces of control information adjusted after the fact.

In addition, the control learning data includes the p pieces of feature information as data corresponding to observation features or feature vectors for artificial intelligence learning, and is included in the observation result or label including the feedback information. It may contain j(i≥1) corresponding data.

Figure 2 is a flowchart showing a process of processing time series data confirmed to have a periodicity that can match the specified periodicity related to the process into frequency domain-based time series data according to the implementation method of the present invention.

In more detail, Figure 2 shows time series data collected and modeled by the operation server 100 through sensors or cameras installed at the service site, and a periodicity that can be matched with a specified periodicity related to at least one process performed at the site. This shows the process of confirming time series data with and processing the time series data into frequency domain-based time series data. Those skilled in the art will refer to and/or modify Figure 2. Therefore, it is possible to infer various implementation methods for the above process (for example, implementation methods in which some steps are omitted or the order is changed), but the present invention includes all of the above inferred implementation methods, and is shown in Figure 2. The technical features are not limited only to the illustrated implementation method.

Referring to Figure 2, the operation server 100 includes S sensing data sensed in time series through S (S ≥ 1) sensors provided at the service site and C (C ≥ 1) cameras provided at the site. Among the C pieces of image data captured in time series, designated D (1≤D≤(S+C)) pieces of collected data are collected (200).

Here, the preprocessing may include an averaging procedure to reveal periodic patterns or representativeness of a certain time interval of the D collected data with a specified periodicity.

In addition, the preprocessing may include a procedure for correcting the temporal influence corresponding to the bias weighted by data of the previous time with respect to the data corresponding to the designated independent variable among the D collected data.

When the D pieces of collected data are collected, the operation server 100 can preprocess the collected D data and match them within the allowable range with T (T ≥ 1) time series models related to the time series of the process performed in the field. Construct N (1≤N≤D) time series data modeled in a similar manner (205).

Here, the T time series models may include a time series model that can match the time series characteristics of a process performed through a device provided at a service site.

Additionally, the N pieces of time series data may include time domain-based time series data based on the time series characteristics of collected data generated by time series sensing or time series photography.

When the N time series data is configured, the operation server 100 has a periodicity that can match a specified periodicity related to at least one process performed in the field based on the N time series data (1≤t≤T ) time series models and n (1≤n≤N) time series data with periodicity that can be matched within the specified allowable range (210).

Here, the periodicity is a periodicity corresponding to at least one process repeatedly performed at regular intervals through a device provided at the service site, and a schedule corresponding to at least one process to be performed through a device provided at the service site. Periodicity, which corresponds to changes in daily intervals (or the interval between Earth's rotation), periodicity that corresponds to changes in trends over a given period, periodicity that corresponds to seasonal changes, and periodicity that corresponds to changes in one-year intervals (or the interval between Earth's revolutions). May contain at least one periodicity.

In addition, the t time series model includes a periodicity that matches the time series characteristics of the process performed in the field and includes multiple periodicities that match the periodic characteristics of the process performed including a specified period. It can be included.

According to the implementation method of the present invention, the operation server 100 analyzes the N time series data based on at least one periodic time series model and n with a periodicity that can match at least one specified periodic time series model within a specified allowable range. You can check time series data.

In addition, according to the implementation method of the present invention, in the case of time series data in which the data change according to time is slower than a specified reference value, the operation server 100 applies a specified exponential smoothing method to the time series data and then generates at least one periodic time series. You can check n time series data with periodicity that can match the model and within the specified tolerance range.

When the n time series data are confirmed, the operation server 100 processes the confirmed n time series data into frequency domain-based time series data corresponding to the specified frequency characteristic (215).

Here, the frequency feature includes a feature of a model-based period (or frequency) corresponding to a time series model with periodicity, or a data-based period (or frequency) confirmed by reading time series data with periodicity, Or, the characteristics of the model-based cycle (or frequency) corresponding to a time series model with periodicity and the characteristics of the data-based cycle (or frequency) confirmed by reading the time series data with periodicity ( or frequency) characteristics.

In addition, the n pieces of time series data processed into the frequency domain are time series data processed into the frequency domain corresponding to the periodicity that matches the periodic characteristics of the process performed, including the period designated at the site, and the frequency characteristics that can be matched within the allowable range. may include.

Figure 3 is a flowchart showing the process of checking analysis information and prediction information using an artificial intelligence module learned according to the implementation method of the present invention.

In more detail, Figure 3 shows that after the process of Figure 2, a plurality of analysis learning data for each target specified in the operation server 100 is configured to learn the artificial intelligence module for analysis, and the analysis is analyzed through the artificial intelligence module for analysis. After configuring a plurality of prediction learning data containing information to train a designated prediction artificial intelligence module, analyze the time series data for each designated target among the time series data collected and preprocessed through the process shown in Figure 2 and processed into the frequency domain. It illustrates the process of confirming the analysis information analyzed by substituting it into an artificial intelligence module for use, and confirming the predicted prediction information by substituting the analysis data including the analysis information into a designated artificial intelligence module for prediction, which is the technology to which the present invention belongs. Those skilled in the art will be able to infer various implementation methods for the above process (e.g., implementation methods in which some steps are omitted or the order is changed) by referring to and/or modifying Figure 3. The present invention includes all of the above-described implementation methods, and its technical features are not limited to only the implementation method shown in Figure 3.

Referring to Figure 3, the operation server 100 processes time series data into frequency domain-based time series data corresponding to a specified frequency characteristic through the process of Figure 2, and then processes n into the frequency domain. Among the time series data, N'(1≤N'≤N, n) for M targets containing n'(1≤n'≤n) time series data for M(M≥1) targets required for analysis of a specified target. Construct a plurality of analysis learning data for each M object containing '⊂N') time series data in a specified structure (300).

Here, the n' time series data are designated among the n time series data processed into the frequency domain corresponding to the periodicity that matches the periodic characteristics of the process performed, including the period specified in the field, and the frequency characteristics that can be matched within the allowable range. It may include at least one time series data necessary for analysis of the target.

In addition, the N' time series data includes n' time series data required for a specified analysis among the n time series data processed into the frequency domain, and i (1≤i) required for a specified analysis among the (N-n) time series data. It may contain ≤(N-n)) time series data.

Here, the operation server 100 can preprocess the N' time series data so that they can be used as learning data for the designated M artificial intelligence modules.

When the plurality of analysis learning data for each of the M objects is configured, the operation server 100 trains the M artificial intelligence modules for analysis using the plurality of analysis learning data for each of the configured M objects (305).

Here, the analysis learning data includes the n' time series data as data corresponding to observation features or feature vectors for artificial intelligence learning, and the observation results or labels related to the n' time series data ( It may contain j(i≥1) pieces of data corresponding to Label).

In addition, the analysis learning data includes the n' time series data, and (n'+) including i(1≤i≤(N-n)) time series data required for analysis of a designated target among the (N-n) time series data. Includes i) time series data as data corresponding to observation features or feature vectors for artificial intelligence learning, and displays observation results or labels (Label) related to the specified time series data among the (n'+i) time series data. ) may include j(i≥1) pieces of data corresponding to ).

After learning the M artificial intelligence modules for analysis, the operation server 100 generates m artificial intelligence modules for analysis corresponding to the designated m (1≤m≤M) objects among the M artificial intelligence modules for analysis. Construct a plurality of prediction learning data containing analysis data including m analysis information analyzed through a specified structure (310), and train the designated artificial intelligence module for prediction using the plurality of prediction learning data constructed (310). 315).

Here, the analysis data includes m analysis information analyzed through the m analysis artificial intelligence modules, and among the n time series data, n" (1≤n"≤n) time series data required for the prediction. may include.

In addition, the predicted learning data includes the analysis data as data corresponding to observation features or feature vectors for artificial intelligence learning, and includes the observation results or labels predicted based on the analysis data. It may contain j(i≥1) corresponding data.

After learning the M artificial intelligence modules for analysis and the artificial intelligence module for prediction, the operation server 100 collects and preprocesses through the process of Figure 2, and then processes the designated M number of time series data into the wave number domain. The M analysis information analyzed is confirmed by substituting the N' time series data for each M object, which includes n' time series data for each object, into the designated M analysis artificial intelligence modules (320).

Here, the operation server 100 may preprocess the N' time series data for each of the M objects so that they can be substituted into the learned M artificial intelligence modules.

When the analyzed M pieces of analysis information are confirmed, the operation server 100 substitutes the analysis data including the designated m pieces of analysis information among the confirmed M pieces of analysis information into the designated prediction artificial intelligence module to obtain predicted prediction information. Check (325).

Thereafter, the operation server 100 stores the confirmed M pieces of analysis information and prediction information in a designated storage medium and provides them to a designated user terminal 140 related to the site (330).

Figure 4 shows that after learning the control artificial intelligence module for controlling the device installed in the field according to the implementation method of the present invention, the characteristic information is substituted into the control artificial intelligence module to generate control information for controlling the device installed in the field. This is a flowchart showing the process of checking and controlling devices in the field.

In more detail, Figure 4 shows the carpenter's control information including control information of devices provided at the service site related to characteristic information including prediction information and designated analysis information confirmed by the operation server 100 after the process of Figure 3. Control learning data is configured to train the control artificial intelligence module for controlling the device, and the designated target is analyzed and the predicted characteristic information is substituted into the control artificial intelligence module to provide control information for controlling the device installed in the field. This shows the process of controlling the device provided in the field, and those skilled in the art can refer to and/or modify Figure 4 to determine various implementation methods for the process. (For example, an implementation method in which some steps are omitted or the order is changed) may be inferred, but the present invention includes all of the above inferred implementation methods, and its technical features are limited to the implementation method shown in Figure 4. This is not limited.

Referring to Figure 4, when M pieces of analysis information and prediction information are confirmed through the process of Figure 3, the operation server 100 provides the confirmed prediction information and designated m' (1≤m'≤M) pieces of analysis information. Contains p (p≥1) pieces of characteristic information, and controls designated e (1≤e≤E) devices related to the feature information among E (E≥1) devices provided at the service site. A plurality of control learning data including the set e pieces of control information are configured (400).

Here, the operation server 100 provides p pieces of feature information including the confirmed prediction information and m' pieces of analysis information to a designated expert terminal, and provides the feature information among the E devices installed at the service site by the expert. You can check the e control information set to control the specified e devices related to.

When the plurality of control learning data is configured, the operation server 100 learns a control artificial intelligence module for controlling the device installed in the field (405).

Here, the operation server 100 can control e devices provided at the site using the e control information, and controls the e devices provided at the site using the e control information. In this case, the actual measurement information for m objects measured currently or after a certain period of time is confirmed for the combination of p feature information and e control information related to the control of the e devices provided in the field, and the p pieces of feature information and By comparing and analyzing the confirmed actual measurement information for each target, feedback information corresponding to control information for improving efficiency related to control of the e devices is generated, including the p feature information and the generated feedback information. The control artificial intelligence module can be additionally trained by configuring control learning data.

After learning the control artificial intelligence module, the operation server 100 analyzes the designated target through the process of Figures 2 and 3 and confirms the predicted p feature information (410).

Then, the operation server 100 substitutes the p feature information into the control artificial intelligence module and confirms the e control information for controlling the e devices installed in the field (415).

When the e pieces of control information are confirmed, the operation server 100 controls the e devices installed at the site using the e pieces of control information (420).

When the e devices installed at the site are controlled using the e control information confirmed through the control artificial intelligence module, the operation server 100 provides p characteristics related to the control of the e devices provided at the site. Regarding the combination of information and e pieces of control information, actual measurement information for each m object measured now or after a certain period of time is checked (425).

Here, the actual measurement information includes actual measurement information for at least one of the following: production volume, production quality, abnormality, failure, and environmental damage corresponding to the results of controlling e devices using the e control information. can do.

In addition, the operation server 100 compares and analyzes the p feature information and the confirmed m object-specific actual measurement information to generate feedback information corresponding to control information for improving efficiency related to control of the e devices. (430).

Here, the feedback information is designated e' (1≤e'≤e) pieces of control information among the e pieces of control information that controlled the e devices provided in the field to improve efficiency related to the control of the e devices. It may include 'e' pieces of control information adjusted after the fact.

When the feedback information is generated, the operation server 100 configures control learning data including the p pieces of feature information and the generated feedback information to further train the control artificial intelligence module (435).

Here, the control learning data includes the p pieces of feature information as data corresponding to observation features or feature vectors for artificial intelligence learning, and is included in the observation result or label including the feedback information. It may contain j(i≥1) corresponding data.

100: Operation server 105: Data configuration unit
110: data confirmation unit 115: data processing unit
120: Artificial intelligence learning unit 125: Information confirmation unit
130: Control module learning unit 135: Control information confirmation unit
140: device control unit 145: device
150: user terminal

Claims (31)

In the method executed through the operating server,
Among the S sensing data sensed in time series through S (S ≥ 1) sensors installed at the service site and C video data captured in time series through C (C ≥ 1) cameras installed at the service site. Designated D (1≤D≤(S+C)) collected data are collected and preprocessed, and N is modeled to enable matching within the allowable range with T (T≥1) time series models related to the time series of the process performed at the site. A first step of constructing (1≤N≤D) time series data;
Based on the N time series data, t (1 ≤ t ≤ T) time series models with a periodicity that can be matched with a specified periodicity related to at least one process performed in the field and n (with a periodicity that can be matched within a specified tolerance range) A second step of checking 1≤n≤N) time series data;
A third step of processing the confirmed n time series data into frequency domain-based time series data corresponding to specified frequency characteristics;
Among the n time series data processed into the frequency domain, M target stars N' (1 ≤N'≤N, n'⊂N') multiple analysis learning data for each target containing time series data in a specified structure are configured to train M artificial intelligence modules for analysis, and the M artificial intelligence modules for analysis are Configures a plurality of predictive learning data containing analysis data containing m analysis information analyzed through m analysis artificial intelligence modules corresponding to m (1≤m≤M) specified targets among the intelligence modules in a specified structure. A fourth step of learning the designated artificial intelligence module for prediction;
After learning the M artificial intelligence modules for analysis and artificial intelligence modules for prediction, designated M targets among the n time series data collected and preprocessed through the processes corresponding to the first to third steps and processed into the frequency domain. Confirm M analysis information analyzed by substituting M objects containing time series data for n' stars into the designated M analysis artificial intelligence modules, and specify m among the confirmed M analysis information. A fifth step of confirming the predicted prediction information by substituting the analysis data including the analysis information into the designated prediction artificial intelligence module;
Among the E(E≥1) devices provided at the service site, it contains p(p≥1) feature information including the confirmed prediction information and specified m'(1≤m'≤M) analysis information. Artificial intelligence for control for controlling devices installed in the field by configuring a plurality of control learning data containing e control information set to control the specified e (1≤e≤E) devices related to the characteristic information A sixth step of learning the module;
After learning of the control artificial intelligence module, when the designated target is analyzed through the process corresponding to the first to fifth steps and the predicted p feature information is confirmed, the p feature information is substituted into the control artificial intelligence module. A seventh step of checking e control information for controlling e devices provided at the site; and
An eighth step of performing a procedure for controlling e devices provided at the site using the e control information. A device control method at a service site comprising a.
The method of claim 1, wherein the pretreatment is,
A device control method at a service site, comprising an averaging procedure to reveal representativeness of periodic patterns or regular time intervals of the D collected data with a specified periodicity.
The method of claim 1, wherein the pretreatment is,
A device control method at a service site, comprising a procedure for correcting a temporal effect corresponding to a bias weighted by data from a previous time with respect to data corresponding to a designated independent variable among the D collected data.
The method of claim 1, wherein the T time series models are:
A device control method at a service site, characterized by including a time series model capable of matching the time series characteristics of a process performed through a device provided at a service site.
The method of claim 1, wherein the N time series data are:
A device control method at a service site comprising time-domain-based time-series data based on time-series characteristics of collected data generated by time-series sensing or time-series photography.
The method of claim 1, wherein the periodicity is:
Periodicity corresponding to at least one process that is repeatedly performed at regular intervals through a device provided at the service site,
Periodicity corresponding to a schedule related to at least one process to be performed through a device provided at the service site,
Periodicity, corresponding to changes in the daily interval (or Earth's rotation interval);
Periodicity, corresponding to changes in trends over a specified period of time;
periodicity in response to seasonal changes;
A device control method at a service site, comprising at least one periodicity corresponding to a change in one-year interval (or Earth orbital interval).
The method of claim 1, wherein the t time series models are:
A service site, characterized in that it includes a periodicity time series model that includes time series matching the time series characteristics of the process performed at the site and includes multiple periodicities matching the periodic characteristics of the process performed including a designated cycle. How to control the device.
The method of claim 1, wherein the second step is:
A service comprising the step of analyzing the N time series data based on at least one periodicity time series model and identifying n time series data with a periodicity that can be matched with at least one specified periodicity time series model and within a specified tolerance range. How to control devices in the field.
The method of claim 1 or 8, wherein the second step is:
In the case of time series data where the data change over time is slower than a specified reference value,
A device control method at a service site comprising the step of applying a specified exponential smoothing method to the time series data and then confirming n time series data with at least one periodicity time series model and a periodicity that can be matched within a specified allowable range.
The method of claim 1 or 8, wherein the second step is:
Analyze the N time series data based on at least one periodic time series model, and separate and analyze data corresponding to random elements or residual/remainder elements included in the time series data to determine at least one periodic time series. A device control method at a service site, comprising the step of checking n time series data with periodicity that can be matched with the model within a specified tolerance range.
The method of claim 1, wherein the frequency characteristic is:
A device control method at a service site, characterized in that it includes features of a model-based period (or frequency) corresponding to a time series model with periodicity.
The method of claim 1, wherein the frequency characteristic is:
A device control method at a service site, characterized by including a data-based period (or frequency) confirmed by reading time series data with periodicity.
The method of claim 1, wherein the frequency characteristic is:
The characteristics of the model-based cycle (or frequency) corresponding to the time series model with periodicity and the characteristics of the data-based cycle (or frequency) identified by reading the time series data with periodicity are statistically processed or numerical analysis is performed to determine the cycle (or A device control method at a service site, characterized in that it includes the characteristics of frequency.
The method of claim 1, wherein the n time series data processed into the frequency domain are:
A device control method at a service site, comprising time series data processed into a frequency domain corresponding to a periodicity that matches the periodic characteristics of a process performed including a specified cycle at the site and a frequency characteristic that can be matched within an allowable range.
The method of claim 1, wherein the n' time series data are:
At least one time series data required for analysis of a designated target among n time series data processed into a frequency domain corresponding to a periodicity that matches the periodic characteristics of a process performed including a specified period in the field and a frequency characteristic that can be matched within an allowable range. A device control method at a service site comprising:
The method of claim 1, wherein the N' time series data are:
Among the n time series data processed into the frequency domain, it includes n' time series data required for analysis of a designated target,
A device control method at a service site, comprising i (1≤i≤(Nn)) time series data required for analysis of a designated target among (Nn) time series data.
The method of claim 1, wherein the fourth step is:
A device control method at a service site, further comprising the step of preprocessing the N' time series data so that they can be used as learning data for designated M artificial intelligence modules.
The method of claim 17, wherein the pretreatment is,
A data change procedure for changing a character or category attribute value for at least one time series data included in the N' time series data into numeric data,
A device control method at a service site, comprising at least one of a data scaling procedure for adjusting a range difference between each time series data included in the N' time series data to within a specified range.
The method of claim 17, wherein the pretreatment is,
A procedure for equalizing the periodicity of at least one time series data included in the N' time series data to a certain time period in a specified time range so that the result value of the next data can be derived through k (k≥1) previous data. A device control method at a service site, comprising:
The method of claim 1, wherein the analysis learning data is:
Containing the n' time series data as data corresponding to observation features or feature vectors for artificial intelligence learning,
A device control method at a service site, comprising j (i≥1) data corresponding to observation results or labels related to the n' time series data.
The method of claim 1, wherein the analysis learning data is:
Artificial intelligence analyzes (n'+i) time series data including the n' time series data and i (1≤i≤(Nn)) time series data required for analysis of a specified target among the (Nn) time series data. Included as data corresponding to observed features or feature vectors for learning,
A device control method at a service site, comprising j(i≥1) data corresponding to an observation result or label related to the specified time series data among the (n'+i) time series data.
The method of claim 1, wherein the analysis data is:
Contains m analysis information analyzed through the m analysis artificial intelligence modules,
A device control method at a service site, comprising n"(1≤n"≤n) time series data required for the prediction among the n time series data.
The method of claim 1, wherein the predicted learning data is:
The analysis data is included as data corresponding to observation features or feature vectors for artificial intelligence learning,
A device control method at a service site, comprising j (i≥1) pieces of data corresponding to observation results or labels predicted based on the analysis data.
The method of claim 1, wherein the fifth step is:
A device control method at a service site, further comprising the step of preprocessing the N' time series data for each of the M objects so that they can be substituted into the learned M artificial intelligence modules.
The method of claim 1, wherein the characteristic information is:
Feature information corresponding to the results of artificial intelligence-based analysis and prediction using frequency domain-based time series data,
A device control method at a service site comprising feature information corresponding to the results of artificial intelligence-based analysis and prediction using multiple frequency domain-based time series data and time domain-based time series data.
The method of claim 1, wherein the sixth step is:
A device control method at a service site, further comprising performing a procedure for controlling e devices provided at the site using the e control information.
The method of claim 1, wherein the sixth step is:
When the e devices provided at the site are controlled using the e control information,
Confirming actual measurement information for m objects measured currently or after a certain period of time with respect to a combination of p feature information and e control information related to the control of the e devices provided in the field;
generating feedback information corresponding to control information for improving efficiency related to control of the e devices by comparing and analyzing the p feature information and the confirmed actual measurement information for each target; and
A device control method at a service site further comprising: configuring control learning data including the p pieces of feature information and the generated feedback information to further train the control artificial intelligence module.
The method of claim 27, wherein the actual measurement information is:
A service site comprising actual measurement information on at least one of production volume, production quality, abnormality, failure, and environmental damage corresponding to the results of controlling the e devices using the e control information. Device control method.
The method of claim 27, wherein the feedback information is:
In order to improve efficiency related to the control of the e devices, e'(1≤e'≤e) designated control information among the e control information that controlled the e devices provided at the site was adjusted ex post. A device control method at a service site, characterized in that it includes control information.
The method of claim 1, wherein the control learning data is:
The p pieces of feature information are included as data corresponding to observation features or feature vectors for artificial intelligence learning,
A device control method at a service site, characterized in that it includes j (i ≥ 1) pieces of data corresponding to observation results or labels including the e pieces of control information.
The method of claim 1, wherein the eighth step is:
When the e devices installed at the site are controlled using the e control information confirmed through the artificial intelligence module for control,
Confirming actual measurement information for m objects measured currently or after a certain period of time with respect to a combination of p feature information and e control information related to the control of the e devices provided in the field;
generating feedback information corresponding to control information for improving efficiency related to control of the e devices by comparing and analyzing the p feature information and the confirmed actual measurement information for each target; and
A device control method at a service site further comprising: configuring control learning data including the p pieces of feature information and the generated feedback information to further train the control artificial intelligence module.