KR20230120763A - System for artificial intelligence-based detection of abnormal signs of cargo anchorage devices and industrial structure safety abnormalities pre-prediction and method thereof - Google Patents

Abstract

The present invention relates to an artificial intelligence-based advance prediction system and method for cargo securing and industrial structure safety anomalies, and more particularly, it is possible to predict cargo securing conditions and safety abnormal signs of industrial structures in advance, and in addition to this, AI-based cargo securing and industrial structures that can quickly respond to accidents that may occur due to deterioration or failure of securing devices by providing information on anomalies and first responders to first responders in the event of safety anomaly signs It relates to a system for predicting safety anomalies in advance and a method therefor.

Description

System for artificial intelligence-based detection of abnormal signs of cargo anchorage devices and industrial structure safety abnormalities pre-prediction and method thereof}

The present invention relates to an AI-based system and method for predicting safety anomalies in cargo securing and industrial structures based on artificial intelligence, and more particularly, an artificial intelligence capable of foreseeing safety anomalies in cargo securing devices and industrial structures in advance. It relates to a system and method for predicting safety anomalies of cargo securing and industrial structures based on intelligence.

In general, in the case of cargo transportation means such as trucks, ships, and aircraft, various types of cargo are loaded and transported by fixing the cargo with a securing device.

This securing device is easy to install and dismantle when fixing in order to prevent the cargo from escaping while the cargo is loaded, the fixing force can be adjusted by adjusting the tension, and the cargo flows together within the range of not escaping. A band-type lashing strap or a wire-type lashing wire is used to minimize damage to cargo by maintaining fixing force while doing so.

This securing device installs a tension adjusting device that adjusts the tension in the state where the cargo is caught in the loading position where the cargo is installed and is fixed in the fixed position, thereby generating a fixing force by adjusting the tension, and adjusting the generated fixing force to the tension. It is installed to secure the cargo held by the device.

However, since the securing device is installed on the means of transportation where the cargo is moved, shaking and external shocks during movement are concentrated on the tension adjusting device that maintains the fixing force, so the securing band is loosened or loosened and the cargo is released from the secured position. or there was a problem with falling to the outside.

Republic of Korea Patent No. 10-1902222 (2018.09.19.)

An object of the present invention is to solve the above-described problems, and an artificial intelligence-based cargo securing device capable of foreseeing safety abnormal signs of industrial structures and cargo securing devices for securing loaded cargo and safety abnormalities of industrial structures It is to provide a symptom advance prediction system and method.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned will be clearly understood by those skilled in the art from the following description.

In order to achieve the above object, the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention is a system for predicting safety anomalies of cargo securing and industrial structures using artificial intelligence technology, A plurality of binding detection devices that are detachable from the binding device to measure the tension of the binding device and wirelessly transmit the measured sensing data using wireless communication technology, and receive the sensing data from the binding detection device and transmit the sensing data to a data management unit A gateway device, a data management unit that receives the sensing data measured by the high-voltage sensing device through the gateway device and stores it in a database server, detects abnormal symptoms of the high-voltage sensing device through machine learning based on the received sensing data, An abnormal symptom diagnosis unit that predicts abnormal symptoms through classification of abnormal symptom types, an abnormal symptom warning unit that determines the severity according to the abnormal symptom type and provides an alarm service according to the determined severity, and the abnormal symptom information and the abnormality of the trouble detecting device and a result providing unit providing result information including first response plan information for symptoms.

In addition, the anomaly symptom diagnosis unit is characterized in that it predicts one of the predefined anomaly symptom types based on the sensing data using supervised learning.

In addition, the high tension sensing device is characterized in that it is configured to further measure vibration, pressure, and temperature.

In addition, the anomaly diagnosis unit detects safety anomalies of industrial structures around the high-voltage detection device through machine learning based on the sensing data including the vibration, pressure, and temperature, and classifies the anomaly symptoms of industrial structures. It is characterized by predicting safety anomalies.

On the other hand, the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction method according to the present invention uses sensing data obtained by measuring state information including tension, ambient temperature, pressure, etc. of the securing device from a plurality of securing devices. Collecting sensing data, detecting abnormal symptoms through machine learning based on the sensing data, and diagnosing abnormal symptoms by predicting abnormal symptoms through classification of abnormal symptom types, determining severity according to the predicted abnormal symptom type and providing an alarm service informing an alarm so that the safety manager or manager can immediately respond according to the determined severity, and the result information including the predicted abnormal symptom information and the initial response plan information for the abnormal symptom is transmitted to the safety manager or It includes a step of providing results to the manager in charge.

Artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system and method according to the present invention can predict safety abnormal signs of cargo securing and industrial structures in advance, By providing abnormal information and initial response plan information, there is an effect of rapidly responding to an accident that may occur due to deterioration or failure of the locking device.

In addition, the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system and method according to the present invention, when safety anomaly signs are predicted, immediately informs each responding organization linked to the securing device and cargo of the accident and the extent of damage spread By being configured to propagate, there is an effect that can solve the spread of large-scale accidents in accidents that may occur from abnormality of the securing device.

1 is a block diagram showing an artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention.
2 is a block diagram showing a data management unit constituting an artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention.
3 is a block diagram showing a database server constituting a data management unit according to the present invention.
4 is an exemplary diagram showing result information of a result providing unit constituting an artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention.
5 is a flow chart showing the overall flow of the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction method according to the present invention.

Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the detailed description of the following embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, only these embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs. It is provided to fully inform the holder of the scope of the invention, and the present invention is only defined by the scope of the claims.

With reference to the accompanying drawings below, specific details for the practice of the present invention will be described in detail. Like reference numbers refer to like elements, regardless of drawing, and "and/or" includes each and every combination of one or more of the recited items.

Terms used in this specification are for describing the embodiments and are not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase. As used herein, "comprises" and/or "comprising" does not exclude the presence or addition of one or more other elements other than the recited elements.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification may be used in a meaning commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless explicitly specifically defined.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram showing an artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention.

Referring to FIG. 1, the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention is a system for predicting safety anomalies of cargo securing and industrial structures using artificial intelligence technology, , a high voltage detection device 100, a gateway device 200, a data management unit 300, an abnormal symptom diagnosis unit 400, an abnormal symptom warning unit 500, and a result providing unit 600.

First, the holddown detection device 100 is attached to and detached from the holddown device to measure state information including the tension state of the holddown device, and performs a function of wirelessly transmitting the measured sensing data using wireless communication technology. dogs can be provided.

To describe the anchorage detection device 100 in more detail, the anchorage detection device 100 is attached to and detached from a anchoring device (an anchoring strap or lashing wire) having one end fixedly coupled to the floor and the other end fixedly coupled to a cargo. It is configured to be detachable from the securing device.

In addition, the fixation detection device 100 includes not only a tension sensor for measuring the tension state of the fixing device coupled to both sides, but also a vibration sensor, a pressure sensor, a temperature sensor, etc. for measuring the surrounding vibration, pressure, and temperature. It may be configured to include, but is not limited thereto. For example, a fire or flooding may be detected by further including a flow rate sensor and a gas sensor.

In addition, the plurality of anchor detection devices 100 each have location information (eg, the location of the cargo deck) for the location of the cargo.

Next, the gateway device 200 serves as a relay between the high tension sensing device 100 and the data management unit 300 so that the sensing data can be received from the high ground sensing device 100 and transmitted to the data management unit 300.

More specifically, the gateway device 200 communicates with the high-stress detection device 100 using wireless communication and communicates with the data management unit 300 through Ethernet communication, and is generally Internet of things ( It means a gateway device that can act as a relay between nodes and hosts for Internet of Things (IoT) communication.

Next, the data management unit 300 plays a role of receiving the sensing data measured by the high-stress detection device 100 through the gateway device 200 and storing it in a database server.

2 is a block diagram showing a data management unit constituting an artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention.

Referring to FIG. 2 , to describe the data management unit 300 in more detail, the data management unit 300 includes a data processing server 310 and a database server 320 .

The data processing server 310 serves to collect sensing data from a plurality of high-failure detection devices 100 .

For example, the data processing server 310 simultaneously receives and collects sensing data while operating a plurality of channels from a plurality of hit detection devices 100, or receives analog sensing data from a plurality of hit detection devices 100 to It is configured to perform a role of various data processing including conversion into digital sensing data, which is only one example and is not limited thereto.

Next, the database server 320 receives and stores the sensing data measured by the fixation detection device, and also serves to store and manage various data used in the fixation device abnormal symptom detection monitoring system according to the present invention. do.

3 is a block diagram showing a database server constituting a data management unit according to the present invention.

More specifically, referring to FIG. 3 , the database server 320 includes a sensor database 321 for storing sensing data measured by the high-voltage detection device, and an abnormal symptom type predicted based on the sensing data, which will be described later. It includes an abnormal symptom type database 322 in which this is stored and an abnormal symptom history database 323 in which the abnormal symptom history diagnosed through the abnormal symptom diagnosis unit 400 is stored.

In addition, the database server 320 further includes a loaded cargo database 324 in which cargo loading plan information including the state and location of cargo is stored, but is not limited thereto.

Next, the abnormal symptom diagnosis unit 400 detects an abnormal symptom of the locking device through machine learning based on the received sensing data, and predicts the abnormal symptom through classification of an abnormal symptom type.

For example, assuming that the predefined abnormal symptom type is normal when the tension of the securing device is 1000 daN, loosened when it is 950 daN, warned when it is 900 daN, and loosened when it is less than 800 daN, the measured value of the specific fixation detection device 100 When the tension is 950 daN (sensing data), the abnormal symptom diagnosis unit 400 detects an abnormal symptom of the fastening device, analyzes it with a predefined abnormal symptom type, and predicts a loosened type. This is only an example for explanation, but is not limited thereto.

In addition, it goes without saying that various abnormal symptoms can be predicted through sensing data including vibration, pressure, temperature, and the like. Here, the various abnormal signs mentioned can be interpreted as meaning including abnormal signs of safety of industrial structures as well as abnormal signs of securing devices.

For the various types of pre-stored anomalies, for example, when the tension is very low, the vibration is very high, the pressure is normal, the temperature is normal, various types such as the type in which the cargo shakes can be learned (stored), and the measured It is expressed as very high, high, normal, low, very low, etc. according to the range of numerical values, and can be set for the type of abnormal symptom for each situation. This is only an example for explanation, but is not limited thereto.

More specifically, the abnormal symptom diagnosis unit 400 predicts one of the predefined abnormal symptom types based on the sensing data using supervised learning.

That is, the type of anomaly is predicted through classification that predicts one of several predefined class labels, which is one of the methods of learning a computer in a state where an explicit answer is given, that is, a label for data.

Preferably, one of the anomaly symptom types is predicted through a decision tree model. The decision tree model is one of the easy methods (algorithms) when dealing with a dataset with many outliers, and is easy to predict one of the types of anomalies through sensing data including tension, vibration, pressure, temperature, and the like.

The decision tree model used in the anomaly symptom diagnosis unit 400 is a method of acquiring information as the purity of a region increases and impurity and uncertainty decrease each time it is separated by the Nth order, and at this time, when calculating uncertainty, Use the coefficient (Gini Index). Here, the reason why the Gini coefficient is used is that log calculation is not required, so the calculation is faster than indicators that quantify impurity such as entropy, so it is suitable for predicting safety anomalies in real time.

On the other hand, the anomaly diagnosis unit 400 uses logistic regression, K-Nearest Neighbors (KNN) in addition to decision trees in predicting safety anomalies of securing devices and industrial structures. ), Naive Bayes, Support Vector Machine (SVM), Random Forest, and Neural Network can be changed and applied in a variety of supervised learning (classification) methods Of course there is.

Next, the abnormal symptom warning unit 500 determines the severity according to the abnormal symptom type determined by the abnormal symptom diagnosis unit 400 and serves to provide an alarm service according to the determined severity.

For example, the severity of the determined abnormal symptoms is determined as normal, interest, caution, vigilance, serious, etc. according to the type of abnormal symptoms such as slight loosening, excessive loosening, and loosening of the securing device, and the alarm service is provided according to the determined severity. to provide.

The warning service may be provided in a form in which a warning lamp or the like is provided at a cargo deck or where a safety manager or a cargo manager works to give various warnings through sounds and colors, but is not limited thereto.

Next, the result providing unit 600 provides result information including abnormal symptom information of the specific failure detection device 100 and initial countermeasure information for the abnormal symptom. This may provide the result information to the terminal side of the safety manager or cargo manager.

Here, the terminal means all kinds of hardware devices including at least one processor capable of visually checking the result information, and according to the embodiment, it will be understood as encompassing a software configuration operating in the corresponding hardware device. can For example, the terminal may be understood as including all user clients and applications running on smartphones, tablet PCs, desktops, laptops, and each device, but is not limited thereto.

In addition, the result providing unit 600 based on the cargo loading plan information including the state and position of the cargo loaded in the cargo transport means stored in the cargo database 324, the result information and the state and It is configured to provide location information together.

4 is an exemplary diagram showing result information of a result providing unit constituting an artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention.

Therefore, as shown in FIG. 4, the result providing unit 600 provides result information including severity according to abnormal symptoms of the hit detection device 100, cargo location and state information, and surrounding state information. can

On the other hand, the AI-based cargo securing and industrial structure safety anomaly prediction system according to the present invention may further include a situation propagation unit 700 as shown in FIG. 1 .

The situation propagation unit 700 serves to deliver the result information and the predicted damage spreading range information to the cargo loading location where the abnormal symptom is detected and the response organization associated with the loaded cargo. Therefore, by disseminating the contents of the accident to surrounding response organizations, there is an effect of preventing an initial response and a problem in which damage may increase in the event of an accident.

In addition, the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention may further include a machine learning learning unit 800.

The machine learning learning unit 800 extracts a major variable based on the stored sensing data and anomaly history, evaluates it to extract a new anomaly symptom type, and performs machine learning on the new anomaly symptom type. do Therefore, there is an effect of improving accuracy and securing reliability in determining an anomaly in machine learning.

In the following, an artificial intelligence-based cargo securing and industrial structure safety anomaly prediction method according to another aspect of the present invention will be described.

5 is a flow chart showing the overall flow of the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction method according to the present invention.

Referring to FIG. 5, the AI-based cargo securing and industrial structure safety anomaly prediction method using the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system according to the present invention is, first, a plurality of securing A sensing data collection step (S10) of collecting sensing data obtained by measuring state information including the tension of the securing device from the sensing device 100 is performed.

Here, the state information may mean the tension state of the securing device, and may also mean state information of an industrial structure including ambient pressure, temperature, flow rate, and gas (eg, CO).

Next, an anomaly symptom diagnosis step (S20) of detecting anomaly symptoms through machine learning based on the sensing data and predicting anomaly symptoms through classification of anomaly symptom types is performed.

Here, machine learning used in the anomaly symptom diagnosis step (S20) refers to a supervised learning method, and preferably, the anomaly symptom diagnosis step (S20) uses supervised learning to predefined based on the sensing data. It is performed to predict one of the anomaly types.

More specifically, the anomaly diagnosis step (S20) is a label for data, that is, a classification that predicts one of several predefined class labels, which is one of the methods of learning a computer in a state where an explicit answer is given. (Classification) predicts the type of anomaly.

Preferably, in the abnormal symptom diagnosis step (S20), one of the abnormal symptom types is predicted through a decision tree model.

The decision tree model used in the anomaly symptom diagnosis step (S20) is a method of acquiring information as the purity of the region increases and the impurity and uncertainty decrease each time the Nth separation is performed. At this time, when calculating the uncertainty Use the coefficient (Gini Index).

Through this, in the abnormal symptom diagnosis step (S20), various safety abnormal symptoms of the securing device or industrial structure can be predicted through sensing data including tension, vibration, pressure, temperature, and the like.

Next, the severity level is judged according to the type of the predicted abnormality, and an alarm service providing step (S30) is performed in which an alarm is notified so that the safety manager or the person in charge of management can respond immediately according to the determined severity.

In the warning service providing step (S30), the severity is judged as normal, interest, caution, alert, serious, etc. according to the type of anomaly, and the place where cargo is loaded, such as a cargo deck, or where a ship safety manager or cargo manager works It is performed to inform the warning through various methods such as sound and color using the provided warning lamp.

Next, a result providing step (S40) of providing result information including the predicted abnormal symptom information and the initial response plan information for the abnormal symptom to the safety manager or the person in charge of management is performed. This may mean providing the result information to the terminal side of a safety manager or a cargo manager.

In addition, the result providing step (S40) is based on the cargo loading plan information including the state and location of the cargo loaded on the cargo transport means stored in the database server 320, the result information and the cargo state and location information. It is performed to provide together.

Next, a situation propagation step (S50) of delivering the result information and the predicted damage spreading range information to the cargo loading location where the abnormal symptom is detected and the response organization associated with the loaded cargo is performed.

On the other hand, the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction method according to the present invention, after the result providing step (S40), extracts key variables based on the collected sensing data and the stored anomaly history , it is possible to perform a machine learning learning step of extracting a new anomaly symptom type by evaluating this, and learning machine learning for the new anomaly symptom type.

Although the embodiments of the present invention have been described with reference to the above and accompanying drawings, those skilled in the art to which the present invention pertains can implement the present invention in other specific forms without changing the technical spirit or essential features. You will understand that there is Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

100: high tension detection device
200: gateway device
300: data management unit
310: data processing server
320: database server
400: Abnormal symptom diagnosis unit
500: Abnormal symptom warning unit
600: result provision unit
700: situation propagation department
800: machine learning learning unit

Claims (5)

In a system for predicting safety anomalies of cargo securing and industrial structures using artificial intelligence technology,
A plurality of binding detection devices that are attached to and detached from the binding device to measure the tension of the binding device and wirelessly transmit the measured sensing data using wireless communication technology;
a gateway device that receives the sensing data from the high voltage detection device and transmits the sensing data to a data management unit;
a data management unit receiving the sensing data measured by the high-failure detection device through the gateway device and storing the data in a database server;
an abnormal symptom diagnosis unit for detecting an abnormal symptom of the high failure detection device through machine learning based on the received sensing data and predicting an abnormal symptom through classification of an abnormal symptom type;
an anomaly warning unit determining a severity according to the type of the abnormal symptom and providing an alarm service according to the determined severity; and
An artificial intelligence-based cargo securing and industrial structure safety anomaly warning system including a result providing unit that provides result information including information on abnormal signs of the anchoring detection device and initial response plan information for the abnormal signs. According to claim 1,
The abnormal symptom diagnosis unit,
An artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system, characterized in that for predicting one of the predefined anomaly symptom types based on the sensing data using supervised learning. According to claim 1,
The high pressure detection device,
An artificial intelligence-based cargo securing and industrial structure safety anomaly prediction system, characterized in that it is configured to measure vibration, pressure, and temperature more. According to claim 3,
The abnormal symptom diagnosis unit,
Based on the sensing data including the vibration, pressure, and temperature, detecting safety anomalies of industrial structures around the high-voltage detection device through machine learning, and predicting safety anomalies of industrial structures through classification of anomaly symptoms AI-based cargo securing and safety anomaly warning system for industrial structures. In the artificial intelligence-based cargo securing and industrial structure safety anomaly prediction method,
Sensing data collection step of collecting sensing data obtained by measuring state information including tension, ambient temperature, pressure, etc. of the securing device from a plurality of securing device:
an abnormal symptom diagnosis step of detecting an abnormal symptom through machine learning based on the sensing data and predicting the abnormal symptom through classification of an abnormal symptom type;
An alarm service providing step of determining severity according to the type of the predicted anomaly and notifying an alarm so that a safety manager or a manager can respond immediately according to the determined severity; and
A result providing step of providing result information including the predicted abnormal symptom information and initial response plan information for the abnormal symptom to a safety manager or manager; Prognosis method.