KR102159108B1 - Network-calibrated, distributive three dimensional fine dust measurement system and method - Google Patents

Abstract

Disclosed are a distributive three-dimensional (3D) fine dust measurement system capable of network calibration to significantly increase measurement reliability, and a method thereof. According to the present invention, the distributive 3D fine dust measurement system capable of network calibration comprises: a plurality of fixed measurement nodes measuring fine dust concentration through an optical sensor and installed in a predetermined area in a fixed type; a plurality of mounted measurement nodes measuring fine dust concentration through an optical sensor and installed in public transportation moving along a preset route; and an autonomous mobile measurement node measuring fine dust concentration through an optical sensor, linked with a preset main fixed measurement node to perform calibration due to effects of temperature and humidity, and moving to at least one of the plurality of fixed measurement nodes and at least one of the plurality of loaded measurement nodes after the calibration to perform calibration.

Description

Network-calibrated, distributive three dimensional fine dust measurement system and method}

The present invention relates to a distributed three-dimensional fine dust measurement system and method capable of network calibration.

Currently, fine dust monitoring in Korea is performed through measuring devices such as BAM (Beta Attenuation Monitoring)-1020.

BAM-1020 uses the principle that when an electron beam is applied to fine dust, the decrease of the electron beam is given as a function only related to the mass according to Beer-Lambert's law.

= 0.873 그리고 0.986 으로 매우 우수한 성능을 보유하고 있다. According to the Federal Reference Method (FRM) evaluation of the Environmental Protection Agency (EPA), the accuracy of BAM-1020 is reliable in 1-hour and 24-hour measurements, respectively??

= 0.873 and 0.986, which has very good performance.

However, BAM-1020 has the following problems.

First, there is a problem that it is expensive and difficult to maintain because high-level technology is used. Therefore, since it is installed only in a limited place, the problem of measurement reliability has been steadily pointed out due to its low spatial resolution. In Seoul, only one place per ward is operated.

Second, the measurement cycle is long. In fact, it takes about an hour for one measurement, and the concentration of fine dust is highly prone to errors due to environmental factors such as solar irradiation or wind and artificial factors such as rush hours.

Third, it is impossible to measure particles of less than 1 micron, and only measure the mass concentration. Currently, Korea uses the mass concentration standard specified by EPA as the air environment standard, but several studies continue to point out that it is more appropriate to reflect the number concentration for the effect on the human body. In addition, in the case of particles of 1 micron or less, they can be accumulated in the brain by passing through the respiratory capillaries as well as the blood-brain barrier. Considering that the major constituents of fine dust are nitrates, sulfates, and heavy metals, it is essential to measure particles of 1 micron or less because it is a serious threat agent that can accumulate in the body and cause neurological, cardiovascular, immune system diseases, and cancer.

Due to these problems, an inexpensive optical fine dust measuring device has been proposed as an alternative. The optical fine dust measuring device irradiates light with an IR-LED to a constant flowing atmospheric sample, measures the light scattered by the fine dust through a photodiode, and continuously at short time intervals based on the Mie theory. It is a device that calculates the concentration of fine dust.

However, due to the problem of measurement accuracy and calibration of the measurement device, construction of a measurement network has not yet been realized. Since the fine dust concentration is calculated through the scattered light received by the photodiode during a unit time interval, the measured value is greatly affected by the stability of the fine dust movement. The accuracy of the measuring device itself has improved considerably as the calculation algorithm has been improved and the design-type atmospheric circulation device has been installed, mainly by overseas companies. However, since the motion of fine dust is still greatly influenced by the temperature and relative humidity of the measurement atmosphere, a correction method is needed to establish a low-cost fine dust measurement network.

Enabling Large-Scale Urban Air Quality Monitoring with Mobile Sensor Nodes, H. David, 2015, ETH

In order to solve the above problems of the prior art, the present invention proposes a distributed three-dimensional fine dust measurement system and method capable of network calibration that can greatly improve measurement reliability even when a low-cost optical sensor is used.

In order to achieve the above object, according to an embodiment of the present invention, according to an embodiment of the present invention, as a distributed fine dust measurement system, the concentration of fine dust is measured through an optical sensor, and a plurality of fixed types are fixedly installed in a preset area. Node; A plurality of on-board measurement nodes installed in public transport moving along a preset route and measuring the concentration of fine dust through an optical sensor; And measuring the concentration of fine dust through an optical sensor, performing calibration according to the influence of temperature and humidity in conjunction with a preset main fixed measurement node, and after completion of the calibration, at least one of the plurality of fixed measurement nodes and the on-board measurement A distributed fine dust measurement system including an autonomous mobile measurement node that moves to at least one of the nodes to perform calibration is provided.

The main fixed measurement node is disposed adjacent to a measuring device for measuring the mass concentration of fine dust, so as to maintain a calibration reliability of a predetermined threshold or higher.

The plurality of fixed measurement nodes, a plurality of mounted measurement nodes, and autonomously mobile measurement nodes may configure a Message Passing Interface (MPI) network to perform parallel computation.

Each of the plurality of onboard measurement nodes determines whether or not the calibration reliability is less than or equal to a preset threshold, and when its own calibration reliability is less than or equal to a preset threshold, another on-board measurement node or a fixed measurement node whose calibration reliability is more than a preset threshold is If adjacent, calibration can be performed using the measured values of the other mounted measurement nodes or fixed measurement nodes as reference data.

The calibration of the first mounted measurement node whose calibration reliability is less than or equal to a preset threshold may be calibrated when the temperature and humidity are the same as that of the other mounted measurement node or the fixed measurement node.

When the number of on-board measurement nodes or fixed measurement nodes adjacent to the first on-board measurement node is less than or equal to a preset threshold, or when the adjacent time is less than or equal to a preset threshold, the first on-board measurement node receives reference data from the MPI network. Multiple calibrations can be performed using.

When the multiple calibration is not possible, the autonomous mobile measurement node moves to the first mounted measurement node, and the first mounted measurement node can perform calibration using the measured value of the autonomous mobile measurement node as reference data. have.

According to another aspect of the present invention, there is provided a method for measuring a distributed fine dust, comprising: measuring a concentration of fine dust by a plurality of fixed measurement nodes fixedly installed in a preset area through an optical sensor; Measuring the concentration of fine dust through an optical sensor by a plurality of on-board measurement nodes installed in public transport moving along a preset route; Performing calibration according to the influence of temperature and humidity in connection with the pre-set main fixed measurement node by the autonomous mobile measurement node; And performing calibration by moving the autonomously mobile measurement node to at least one of the plurality of fixed measurement nodes and at least one of the mounted measurement nodes after calibration is completed.

According to the present embodiment, first, it is possible to operate a large-scale sensor measurement node network with low unit cost and maintenance cost. This can significantly improve the resolution compared to the existing fine dust measurement network.

In addition, the measurement period is as short as 1 second or less, enabling continuous fine dust monitoring. Therefore, it is possible to significantly improve the measurement cycle of the existing one-hour unit.

Furthermore, it is possible to collect 3D fine dust data by using an autonomous mobile measurement node, and has the advantage of enabling more precise fine dust model research.

A number of measurement nodes are operated in a form mounted on public transport. Therefore, in an urban area with a high density of public transportation, the above two effects are doubled.

Next, since the network is built through MPI, the measurement node cluster can operate a web server for data publication, process its own operation, and calculate a model. Therefore, it has the effect of reducing the cost required for server operation.

Third, the use of optical sensors enables the monitoring of ultrafine dust, which is the most lethal to the human body. Unlike the BAM (Beta Attenuation Monitor) method, the optical fine dust measurement method is also advantageous for the measurement of very small particles.

1 is a view showing a fine dust measurement system according to an embodiment of the present invention.
2 is a diagram showing the configuration of a measurement node according to the present embodiment.
3 is a diagram illustrating an operation process of a fine sensitivity measurement system according to an embodiment of the present invention.
4 is a view showing a calibration process according to the contribution of humidity and temperature according to an embodiment of the present invention.
5 is a diagram illustrating a calibration process of a mounted measurement node according to an embodiment of the present invention.
6 is a diagram illustrating a process of assigning calibration reliability according to an embodiment of the present invention.
7 is a diagram illustrating a calibration process of a measurement node to be calibrated first according to the present embodiment.

In the present invention, various modifications may be made and various embodiments may be provided, and specific embodiments will be illustrated in the drawings and described in detail.

However, this is not intended to limit the present invention to a specific embodiment, it is to be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the present invention.

1 is a view showing a fine dust measurement system according to an embodiment of the present invention.

As shown in FIG. 1, the fine dust measurement system according to the present embodiment may include a plurality of fixed measurement nodes 100 and 102, an autonomous mobile measurement node 104 and a mounted measurement node 106.

One of the plurality of fixed measurement nodes is disposed adjacent to a measurement device that measures the mass concentration of fine dust, and maintains a calibration reliability above a preset threshold with the highest priority, and this is defined as the main fixed measurement node 100.

The main fixed measurement node 100 is placed in a measurement station belonging to the existing Environment Corporation, and maintains the latest calibration state by interlocking with a measurement device that measures the mass concentration of fine dust with high reliability, such as BAM-1020 installed at the measurement station belonging to the Environment Corporation. .

In addition, the fixed measurement node 102 installed in the area is disposed in an area of interest with a large amount of population flow and traffic, and an area where there are many intersections of traffic lines.

The fixed measurement nodes 100 and 102 arranged in the region of interest improve the efficiency of propagating the calibration throughout the measurement nodes constituting the network and the monitoring capability of the measurement network.

The autonomous mobile measurement node 104 may be an unmanned aerial vehicle such as a drone, and calibration according to the influence of temperature or humidity is performed using the measured value of the main fixed measurement node 100 that maintains calibration reliability above a preset threshold as reference data. And move to the onboard measurement node 106.

The on-board measurement node 106 according to this embodiment may be installed in public transportation.

The autonomous mobile measurement node 104 moves to a parking lot or a route intersection where the on-board measurement node 106 is concentrated and performs secondary calibration on the on-board measurement node 106. In addition, when abnormal measurement value change is monitored on the network, it has the role of moving directly to further monitor it.

The mounted measurement node 106 constitutes a measurement network that monitors fine dust together with the fixed measurement nodes 100 and 102, and performs calibration of other mounted measurement nodes 106 that are adjacent while moving along a route. This allows a final calibration of all globally deployed on-board measurement nodes to be performed.

The onboard measurement node 106 connects to the network through a data hotspot that has started to be loaded on public transportation and performs calibration of the other onboard measurement node 106. For this, port forwarding is required for the device.

The measurement network and calibration network according to the present embodiment may be configured based on a Message Passing Interface (MPI).

That is, according to the present embodiment, a plurality of fixed measurement nodes 100 and 102, an autonomous mobile measurement node 104 and a mounted measurement node 106 constitute an MPI network.

MPI is a standardized information delivery system and software used for parallel computing. It is involved in the exchange of information between CPU cores and integrates a plurality of computing devices into a single computing system. It is used not only for data parallelism used in supercomputers with high computational speed, but also for task parallelism used for efficient separation of functions.

Since MPI is generally used for tasks that require high performance, microcontrollers that can operate it are limited because it is optimized for major operating systems and CPU architectures. For this reason, the measurement node of this embodiment is based on the Raspberry Pi platform. Raspberry Pi runs on the ARM Architecture core. Compared to other microcontrollers, digital signal input and output expandability is excellent, and it is possible to operate various operating systems as well as Raspbian, a basic operating system based on Debian OS. Therefore, it is possible to maintain all elements of the network with a single platform of Raspberry Pi, and the structure of this network has very high stability regardless of its size.

The fine dust measurement system according to the present embodiment is composed of distributed measurement nodes without a host. Access between measurement nodes is made through SSH (Secure Shell), and it is MPI that connects them with a program.

Each measurement node simultaneously operates a calibration program (MPI for calibration) and a measurement network program (MPI for measurement) through MPI. The measured data is basically directly managed by the corresponding measurement node, and periodically analyzed data is uploaded to the data web server 110.

The calibration MPI is directly involved in the calibration process while checking the calibration status, and the measurement network MPI analyzes the calibrated measurement value in the measurement node or uploads it to the data web server 110.

2 is a diagram showing the configuration of a measurement node according to the present embodiment.

FIG. 2 shows a configuration common to a fixed type, an autonomous mobile type, and a mounted type measuring node.

As shown in FIG. 2, all measurement nodes may include a sensor unit 200, a storage unit 202, a communication unit 204, and a control unit 206.

The sensor unit 200 is an optical sensor, and measures light scattered by fine dust through a photodiode by irradiating light onto an atmospheric sample with an IR-LED.

As described above, the storage unit 202 stores two types of data, an original measurement value through the sensor unit 200 and a measurement value calibrated through another measurement node.

The communication unit 204 transmits the calibrated measurement value to the data web server 110 or to another measurement node.

The control unit 206 controls the operation of the sensor unit 200, the storage unit 202, and the communication unit 204.

EPA's FRM and FEM are representative methods of evaluating the sensor's capabilities, and these are the same methods that can be used to calibrate the sensor. FRM is a method of measuring a sample whose value is known accurately, and FEM is a method of measuring with other sensors that have confirmed high reliability.

The calibration according to this embodiment is similar to the FEM. Specifically, a method of performing the first calibration with BAM-1020 or other verified measurement node with secured reliability and Ordinary least squares (OLS) regression, and then using the calibrated measurement node to fully calibrate other measurement nodes. to be.

The calibration method according to this embodiment is premised on comparing the measured values of each measurement node while having the same measurement conditions. Therefore, it is necessary to measure at the same place at the same time for calibration.

BAM-1020 is suitable for use as standard data for calibration because it has high accuracy. However, since the measurement cycle is very long (1 hour), it is necessary to accumulate data over a long period of time in order to perform OLS regression. For this reason, by arranging the main fixed measurement node 100 at the measurement station of the Environment Corporation, the latest calibration state is always maintained.

After the first calibration, calibration between other measurement nodes and autonomous mobile measurement nodes is possible in a short time. In general, inexpensive optical sensors have a measurement cycle of several seconds, so sufficient data can be accumulated for a short time.

3 is a diagram illustrating an operation process of a fine sensitivity measurement system according to an embodiment of the present invention.

Referring to FIG. 3, the main fixed measurement node 100 disposed at the environmental industrial complex measurement station maintains the latest calibration value (step 300).

The main fixed measurement node 100 according to the present embodiment downloads the data of the environmental agency measurement station (BAM-1020) through the MPI network, performs data calibration of the environmental agency measurement station, and performs calibration every 24 hours or more.

Next, the autonomous mobile measurement node 104 approaches the main stationary measurement node 100 and performs calibration using the measured value of the main stationary measurement node 100 as reference data (step 302).

The calibration is completed, the autonomous mobile measurement node 104 moves to a public transportation hub or garage and performs calibration of the on-board measurement node 106 (step 304).

In addition, the mounted measurement node 106 according to the present embodiment performs calibration based on the calibration reliability whenever it is adjacent to the other mounted measurement node 106 or the fixed measurement node 102, and updates the calibration reliability. (Step 306).

The measurement node on which the calibration is completed periodically uploads the measurement value to the data web server 110 (step 308).

4 is a view showing a calibration process according to the contribution of humidity and temperature according to an embodiment of the present invention.

4 shows a process performed by each measurement node to be calibrated.

Referring to FIG. 4, the measurement node stores the original measurement value (step 400), and then determines whether or not the contribution of the relative humidity is large (step 402).

Step 402 may be determined based on reference data, that is, a measurement value obtained from another measurement node that has been calibrated.

If the relative humidity contribution is large, calibration is performed according to the relative humidity (step 404).

In step 404, in general, for humidity calibration, a Hanel growth function such as Equation 1 below or a Soneja growth function such as Equation 2 may be used.

Here, a and b are coefficients, RH is relative humidity, the original measurement value is the value directly measured by the measurement node, and the reference measurement value is the value stored by the measurement node that has been calibrated.

The humidity correction value is calculated as follows through the humidity correction coefficient calculated through Equations 1 to 2.

It is determined whether the contribution of the relative humidity is not large in step 402, or whether the contribution of temperature is large after calibration according to the relative humidity is completed in step 404 (step 406).

In addition, in the case of temperature calibration, Equation 4 or Equation 5 below is used.

은 절편에 해당하는 계수,

는 기준 측정값에 대한 비례 계수,

은 절대온도에 대한 비례 계수이다. In Equations 4 to 5,

Is the coefficient corresponding to the intercept,

Is the proportionality factor for the reference measurement,

Is the coefficient of proportion to the absolute temperature.

,

,

를 원 측정값에 대입하여 최종된 보정값을 산출하는 식을 나타낸 것이다. Equation 6 below is calculated through Equation 4

,

,

This is an equation for calculating the final correction value by substituting for the original measured value.

In addition, the following Equation 7 shows the result of removing the term due to low temperature contribution.

In step 406, if the influence of temperature is large, calibration is performed according to the temperature correction value of Equation 6 (step 408), and if the influence of temperature is not large, calibration is performed using the temperature correction value of Equation 7. (410).

According to this embodiment, there are two factors that trigger the calibration between measurement nodes. The first is a method of performing p2p calibration by operating MPI with data I/O on the existing MPI network or LAN (Local Area Network) when physically in a short distance.

The other is to perform calibration by synthesizing all data measured at the same time and location on the MPI network.

The first trigger has high stability and can be used in any type of calibration. The second trigger increases calibration opportunities, improves calibration reliability, and is primarily used in on-board measurement node-to-node calibration.

Calibration between on-board measuring nodes or for fixed measuring nodes placed in an area of interest may not satisfy the perfect level of calibration conditions. In particular, this is the case where physical close proximity is the trigger. An example of this is the case where the relative speed between the two measurement nodes is so large that data accumulation within a short distance for which calibration is effective is not sufficiently performed. Therefore, each time calibration is performed, calibration reliability should be given to the measurement node so that the measured value and the reliability of calibration performed through the node should be evaluated.

5 is a diagram illustrating a calibration process of a mounted measurement node according to an embodiment of the present invention.

FIG. 5 shows a process in which a mounted measurement node performs calibration through another adjacent mounted measurement node or a fixed measurement node.

Referring to FIG. 5, the onboard measurement node 106 stores the original measurement value (step 500).

Next, it is determined whether or not the location of the other measurement node is within a threshold (step 502).

When the two measurement nodes are within a predetermined distance, the onboard measurement node 106 determines whether or not the calibration reliability of the measurement node adjacent to it is greater than or equal to a threshold (step 504).

The calibration reliability according to this embodiment can be expressed by the following equation.

The shelf life reliability decreases as time elapses from the last calibration date, and the sample number reliability increases as the number of measurement data used at the time of calibration increases.

When the calibration reliability is greater than or equal to a preset threshold (eg, greater than or equal to 0.8), the onboard measurement node 106 determines whether the temperature and humidity are the same as that of the adjacent measurement node (step 506).

If the temperature and humidity are the same, calibration is performed based on a measurement node having high calibration reliability (step 508).

After the calibration is completed, the calibration reliability of the on-board measurement node 106 is updated, and the calibrated measurement value is stored (step 512).

6 is a diagram illustrating a process of assigning calibration reliability according to an embodiment of the present invention.

6, each measurement node periodically checks its own calibration reliability (step 600).

Each measurement node determines whether its own calibration reliability is 0.8 or more (first threshold) (step 602).

The calibration reliability can be determined according to Equation 8 above, and when the calibration reliability is 0.8 or more, it is used as a calibration standard and measurement node (step 604).

On the other hand, when the calibration reliability is 0.8 or less, new data is accumulated, or when measurement nodes having high calibration reliability are adjacent to each other, recalibration is automatically performed (step 606).

Further, each measurement node determines whether or not the calibration reliability is 0.7 or more (a second threshold) (step 608).

If the calibration reliability is 0.7 or more, it is not used as a calibration standard, but is used only as a measurement node (step 610).

On the other hand, if the calibration reliability is 0.7 or less, the measurement node is first classified as a measurement node to be calibrated (step 612).

In step 610, when it is first determined as a measurement node to be calibrated, the autonomous mobile measurement node 104 is first moved to the measurement node to be calibrated in the MPI network, or when other measurement nodes are adjacent, or through a calibration value download Perform node calibration.

7 is a diagram illustrating a calibration process of a measurement node to be calibrated first according to the present embodiment.

FIG. 7 illustrates a process performed when the number of on-board measurement nodes or fixed measurement nodes adjacent to a measurement node to be calibrated is less than or equal to a preset threshold or the adjacent time is less than or equal to a preset threshold.

Referring to FIG. 7, first, a measurement node to be calibrated searches and downloads a location adjacent to another measurement node, an adjacent time, and a measurement value (step 700).

Step 700 may be first performed by interworking with the MPI network in which the measurement node to be calibrated performs parallel operation according to the present embodiment.

First, the measurement node to be calibrated performs calibration with the measurement values of other measurement nodes downloaded from the MPI network (step 702).

Further, data is accumulated until the calibration reliability becomes 0.8 or more, and calibration is periodically performed (step 704).

As described above, steps 700 to 704 may be performed when the first measurement node to be calibrated does not have many opportunities to be adjacent to other measurement nodes or calibration is not frequently performed.

That is, if it does not have a sufficient opportunity and time adjacent to other measurement nodes, the measurement node to be calibrated first downloads the measurement value from the MPI network and performs calibration.

Steps 700 to 704 may be defined as multiple calibrations.

If calibration cannot be performed even through the above-described process, that is, when multiple calibration is difficult, the autonomously mobile measurement node first moves to the measurement node to be calibrated to perform calibration directly (step 706).

The movement of the autonomous mobile measurement node can be accomplished through parallel computation in the MPI network.

The measurement node according to the present embodiment may have an MPI installed on the Raspberry Pi.

The Raspberry Pi microcontroller is an ARM architecture and runs based on the Rapbian operating system built on Debian. However, Rapbian is running the system by downgrading the CPU to armv7, a 32-bit based architecture, so that the GNU compiler runs on the operating system. For this reason, in Raspberry Pi, both Mpich and OpenMPI, which are representative MPI implementations, confuse CPU architectures, and memory barrier conflicts may occur on the kernel.

Therefore, the installation process must be modified so that MPI can run normally on the Raspberry Pi, and the CPU architecture must be forcibly assigned to armv8 or higher. In this embodiment, normal operation of P2P communication and collective communication in Mpich and OpenMPI was confirmed through architectural force allocation.

Also, since MPI is controlled through the SSH (Secure Shell) protocol, SSH access between Raspberry Pis must be made, and the Raspberry Pi supports this by default. In this embodiment, an SSH connection between measurement nodes is implemented by accessing the Internet using a WiFi module.

SSH access methods can be considered both LAN and remote methods. In the case of LAN, it can be mainly used for calibration of the proximity trigger method by using WiFi standard communication. The remote method can also consider access through the Internet and access through the WAN. For WAN, you can consider using LoRa (Long Range).

The above-described embodiments of the present invention have been disclosed for the purpose of illustration, and those skilled in the art who have ordinary knowledge of the present invention will be able to make various modifications, changes, and additions within the spirit and scope of the present invention, and such modifications, changes and additions It should be seen as belonging to the following claims.

Claims (10)

As a distributed fine dust measurement system,
A plurality of fixed measurement nodes that measure the concentration of fine dust through an optical sensor and are fixedly installed in a preset area;
A plurality of on-board measurement nodes installed in public transport moving along a preset route and measuring the concentration of fine dust through an optical sensor; And
Measures the concentration of fine dust through an optical sensor, accesses a preset main fixed measurement node, performs calibration according to the influence of temperature and humidity, and after completion of the calibration, at least one of the plurality of fixed measurement nodes and the on-board measurement node Including an autonomous mobile measurement node for performing calibration by moving to at least one of,
The main fixed measurement node is disposed adjacent to a measuring device that measures the mass concentration of fine dust, and maintains calibration reliability above a preset threshold with the highest priority,
The autonomous mobile measurement node is an unmanned aerial vehicle, and a distributed fine dust measurement system for performing calibration by moving to a fixed measurement node or a mounted measurement node having a calibration reliability below a preset threshold. delete The method of claim 1,
The plurality of fixed measurement nodes, a plurality of mounted measurement nodes, and autonomously mobile measurement nodes constitute a MPI (Message Passing Interface) network to perform parallel computation. The method of claim 3,
Each of the plurality of on-board measurement nodes determines whether or not the calibration reliability is less than a preset threshold, and when its calibration reliability is less than a preset threshold, another on-board measurement node or a fixed measurement node whose calibration reliability is more than a preset threshold is If adjacent, a distributed fine dust measurement system that performs calibration using the measured values of the other mounted measurement nodes or fixed measurement nodes as reference data. The method of claim 4,
A distributed fine dust measurement system performed when the calibration reliability of the first mounted measurement node is less than or equal to a preset threshold when the temperature and humidity are the same as that of the other mounted measurement node or the fixed measurement node. The method of claim 5,
When the number of on-board measurement nodes or fixed measurement nodes adjacent to the first on-board measurement node is less than or equal to a preset threshold, or when the adjacent time is less than or equal to a preset threshold, the first on-board measurement node receives reference data from the MPI network. Distributed fine dust measurement system that performs multiple calibrations using. The method of claim 6,
When the multiple calibration is not possible, the autonomous mobile measurement node moves to the first mounted measurement node, and the first mounted measurement node performs calibration using the measured value of the autonomous mobile measurement node as reference data. Type fine dust measurement system. As a distributed fine dust measurement method,
Measuring a concentration of fine dust through an optical sensor by a plurality of fixed measurement nodes fixedly installed in a preset area;
Measuring a concentration of fine dust through an optical sensor by a plurality of on-board measurement nodes installed in public transport moving along a preset route;
Performing calibration according to the influence of temperature and humidity by the autonomous mobile measurement node accessing the preset main fixed measurement node; And
After the calibration is completed, the autonomous mobile measurement node moves to at least one of the plurality of fixed measurement nodes and at least one of the on-board measurement nodes to perform calibration,
The main fixed measurement node is disposed adjacent to a measuring device that measures the mass concentration of fine dust, and maintains calibration reliability above a preset threshold with the highest priority,
The autonomous mobile measurement node is an unmanned aerial vehicle, and a distributed fine dust measurement method for performing calibration by moving to a fixed measurement node or a mounted measurement node having a calibration reliability below a preset threshold. delete The method of claim 8,
The plurality of fixed measurement nodes, a plurality of mounted measurement nodes, and autonomously mobile measurement nodes constitute a Message Passing Interface (MPI) network to perform parallel computation.

Images