Search

KR-102962997-B1 - Apparatus for suppressing dust on construction site based on diagnosis using deep learning model (DLM) and method therefor

KR102962997B1KR 102962997 B1KR102962997 B1KR 102962997B1KR-102962997-B1

Abstract

The method for suppressing dust according to the present invention comprises the steps of: a data processing unit receiving weather information from a weather information server and receiving fine dust information from an environment information server through a communication module; the data processing unit receiving fugitive dust information measured by a sensor at a construction site; a diagnostic unit analyzing the fugitive dust information, the fine dust information, and the weather information through a diagnostic network which is a trained deep learning model to diagnose whether suppression of dust generated at the construction site is required; and, if it is diagnosed that dust suppression is required, a driving unit transmitting an operation command through the communication module to a dust suppression device deployed at the construction site to perform dust suppression.

Inventors

  • 이준성
  • 박영현

Assignees

  • 에스케이플래닛 주식회사

Dates

Publication Date
20260511
Application Date
20210215

Claims (11)

  1. A step in which a data processing unit receives fugitive dust information measured by a sensor at a construction site, receives weather information for the zone to which the construction site belongs from a weather information server via a communication module, and receives fine dust information for the zone to which the construction site belongs from an environment information server; A step of inputting the fugitive dust information, the fine dust information, and the weather information into an input layer comprising a plurality of input nodes of a diagnostic network, which is a deep learning model trained by a diagnostic unit, and performing a plurality of operations to which weights are applied through at least one hidden layer of the diagnostic network to diagnose whether suppression of dust generated at the construction site is required; If it is diagnosed that dust suppression is required, the driving unit transmits an operation command through the communication module to the dust suppression device deployed at the construction site to perform dust suppression; the method includes. Prior to the step of receiving the fine dust information, the diagnostic network performs learning based on learning analysis data including an experimental group in which the dust concentration at the actual construction site is measured in a state where the operation of the dust suppression device is not required, and a control group in which the dust concentration at the actual construction site is measured in a state where the operation of the dust suppression device is required; and The above-mentioned analysis data is, A method for suppressing dust, characterized by including fugitive dust information indicating the concentration of fugitive dust measured at the actual construction site, fine dust information for the area including the actual construction site collected from the environmental information server, and weather information for the area including the actual construction site collected from the weather information server.
  2. ◈Claim 2 was waived upon payment of the establishment registration fee.◈ In paragraph 1, The step of diagnosing whether dust suppression is required at the aforementioned construction site A step of calculating the probability of whether the above diagnostic network is in a state where operation of the dust suppression device is required or where operation of the dust suppression device is not required; and A method for suppressing dust, characterized by including the step of the diagnostic unit diagnosing whether the dust suppression device is operating according to the calculated probability.
  3. ◈Claim 3 was waived upon payment of the establishment registration fee.◈ In paragraph 1, The step of performing the above learning is, A step in which the learning unit sets labels for the above-mentioned learning analysis data using one-hot encoded vectors; A method for suppressing dust, further comprising the step of training the diagnostic network using the above-mentioned labeled analysis data.
  4. ◈Claim 4 was waived upon payment of the establishment registration fee.◈ In paragraph 3, The above-mentioned learning step A step in which the above learning unit inputs the learning analysis data with the above-mentioned label set into the diagnostic network; A step in which the diagnostic network performs multiple operations in which weights between multiple layers are applied to the training analysis data with the label set, and calculates as an output value the probability regarding whether the operation of the dust suppression device is required or not required; A method for suppressing dust, characterized by including the step of updating the weights of the diagnostic network through a backpropagation algorithm so that the loss, which is the difference between the output value and the label, is minimized by the learning unit.
  5. ◈Claim 5 was waived upon payment of the establishment registration fee.◈ In paragraph 4, A step of updating the weights of the above diagnostic network; The above learning unit mathematical formula So that the loss calculated according to is minimized The weights of the diagnostic network are updated through the backpropagation algorithm above, and The above CE is a loss representing the difference between the output value and the label, and The above Lx is a label, and A method for suppressing dust characterized in that the above Sx is an output value.
  6. Communication module for communication; A data processing unit that receives fugitive dust information measured by a sensor at a construction site, receives weather information for the zone to which the construction site belongs from a weather information server via the communication module, and receives fine dust information for the zone to which the construction site belongs from an environment information server; A diagnostic unit that inputs the fugitive dust information, the fine dust information, and the weather information into an input layer including a plurality of input nodes of a diagnostic network, which is a trained deep learning model, and performs a plurality of operations to which weights are applied through at least one hidden layer of the diagnostic network to diagnose whether suppression of dust generated at the construction site is required; A driving unit that transmits an operation command via the communication module to perform dust suppression to a dust suppression device deployed at the construction site when it is diagnosed that dust suppression is required; The above diagnostic network includes a learning unit that performs learning based on learning analysis data comprising an experimental group in which the dust concentration at the actual construction site is measured in a state where the operation of the dust suppression device is not required, and a control group in which the dust concentration at the actual construction site is measured in a state where the operation of the dust suppression device is required; The above-mentioned analysis data is, A device for suppressing dust, characterized by including fugitive dust information indicating the concentration of fugitive dust measured at the actual construction site, fine dust information for the area including the actual construction site collected from the environmental information server, and weather information for the area including the actual construction site collected from the weather information server.
  7. ◈Claim 7 was waived upon payment of the establishment registration fee.◈ In paragraph 6, The above diagnostic network is The above input layer; The at least one hidden layer comprising a plurality of hidden nodes, each including an operation; A device for suppressing dust, characterized by including an output layer comprising two output nodes, each including an operation.
  8. ◈Claim 8 was waived upon payment of the establishment registration fee.◈ In Paragraph 7, The two output nodes above are A first output node that outputs the probability that the dust condition of the above construction site is a state requiring the operation of the dust suppression device, and A device for suppressing dust, characterized by including a second output node that outputs the probability that the dust condition of the construction site is such that the operation of the dust suppression device is unnecessary.
  9. ◈Claim 9 was waived upon payment of the establishment registration fee.◈ In paragraph 6, The above learning unit A device for suppressing dust, characterized by setting labels for the training analysis data using one-hot encoding vectors and training the diagnostic network using the training analysis data with the set labels.
  10. ◈Claim 10 was waived upon payment of the establishment registration fee.◈ In Paragraph 9, The above learning unit Input the above-mentioned training analysis data into the diagnostic network, and If the diagnostic network performs multiple operations with weights applied between multiple layers on the learning analysis data to calculate as an output value the probability regarding whether the operation of the dust suppression device is required or not, A device for suppressing dust, characterized by updating the weights of the diagnostic network through a backpropagation algorithm so that the loss, which is the difference between the output value and the label, is minimized.
  11. ◈Claim 11 was waived upon payment of the establishment registration fee.◈ In Paragraph 10, The above learning unit mathematical formula So that the loss calculated according to is minimized The weights of the diagnostic network are updated through the backpropagation algorithm above, and The above CE is a loss representing the difference between the output value and the label, and The above Lx is a label, and A device for suppressing dust, characterized in that the above Sx is an output value.

Description

Apparatus for suppressing dust on construction site based on diagnosis using deep learning model (DLM) and method therefor The present invention relates to a technology for suppressing dust at a construction site, and more specifically, to an apparatus for suppressing dust at a construction site based on diagnosis using a Deep Learning Model (DLM) and a method for the same. Dust refers to particulate matter that floats or drifts down in the atmosphere, and it is known that the smaller the particle size, the greater the adverse effect on the lungs. Common expressions for atmospheric dust concentration include TSP (Total Suspended Particles), PM-10 (Particulate Matter 10), and PM2.5. Total Suspended Particles (TSP) typically refers to all suspended dust particles smaller than 50 μm. Since particles larger than 10 μm affect urban aesthetics but have minimal impact on human health, the environmental standard was changed from TSP to PM-10 in the late 1990s. Particulate matter (PM10) refers to solid particles and liquid droplets mixed and floating in the air. PM10 has a diameter of 2.5 to 10 µm and is primarily generated from construction and building demolition, coal and oil combustion, industrial processes, and unpaved roads. Fine particles (PM2.5) have a diameter of less than 2.5 µm and are mostly generated through chemical reactions in the atmosphere from the combustion of coal, oil, gasoline, diesel, and wood, as well as from smelters and steel mills. Fugitive dust, also known as fugitive dust, is a general term for dust that is emitted directly into the atmosphere without a designated exhaust outlet. Fugitive dust is frequently generated in industries such as construction, cement, coal, and soil. In particular, fugitive dust emitted from construction sites varies significantly depending on the daily volume of work, construction methods, and weather conditions. If the Minister of Environment determines that facilities for suppressing the generation of fugitive dust have not been installed or necessary measures have not been taken, or that such facilities or measures are deemed inadequate, the Minister may order the person conducting the business to install necessary facilities or implement or improve the measures. FIG. 1 is a drawing illustrating a system for suppressing dust at a construction site based on diagnosis using a deep learning model according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating the configuration of a control server according to an embodiment of the present invention. FIG. 3 is a block diagram illustrating the detailed configuration of a control module of a control server according to an embodiment of the present invention. FIG. 4 is a diagram illustrating the configuration of a diagnostic network, which is a deep learning model according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating a method for training a diagnostic network through deep learning to diagnose the condition of dust at a construction site according to an embodiment of the present invention. FIG. 6 is a flowchart illustrating a method for suppressing dust at a construction site based on diagnosis using a deep learning model according to an embodiment of the present invention. In order to clarify the features and advantages of the means for solving the problem of the present invention, the present invention will be described in more detail with reference to specific embodiments of the present invention illustrated in the attached drawings. However, detailed descriptions of known functions or configurations that may obscure the essence of the invention are omitted in the following description and the attached drawings. Additionally, it should be noted that identical components throughout the drawings are indicated by the same reference numerals whenever possible. Terms and words used in the following description and drawings should not be interpreted as being limited to their ordinary or dictionary meanings, but should be interpreted in a meaning and concept consistent with the technical spirit of the invention, based on the principle that the inventor can appropriately define the concept of terms to best describe his invention. Accordingly, the embodiments described in this specification and the configurations illustrated in the drawings are merely the most preferred embodiments of the invention and do not represent all aspects of the technical spirit of the invention; therefore, it should be understood that various equivalents and modifications capable of replacing them may exist at the time of filing this application. Furthermore, terms including ordinal numbers, such as first, second, etc., are used to describe various components and are used solely for the purpose of distinguishing one component from another, and are not used to limit said components. For example, without departing from the scope of the present invention, the second component may be name