JP2020106916A - Setting method of monitor event analysis device - Google Patents

Abstract

To more appropriately identify a state of a monitored object such as a moving body and a structure, without requiring complicated wiring or an external power supply.SOLUTION: A setting method of a monitor event analysis device identifies a type of an event occurred in a monitored object, and transmits an identification result. The analysis device comprises: sensors which measure a state of the monitoring object; an identification processing unit which executes identification processing to generate an identification result based on an event identification program; and a radio communication unit which transmits the identification result. The setting method of the monitor event analysis device comprises: attaching an information collection device with sensors which measure the state of the monitoring object to the monitoring object; inputting sensor information by the information collection device; inputting event identification information corresponding to each type of the event; transmitting time sequence information having the sensor information and the event identification information to a monitor information processing device as learning data; executing machine learning using the learning data; generating an event identification program which identifies the type of the event based on the sensor information; and installing the program in the analysis device.SELECTED DRAWING: Figure 4

Description

The present invention relates to a device for monitoring a monitoring target such as a moving body or a structure, and particularly to a setting method of a monitoring event analysis device attached to the monitoring target.

In recent years, the maintenance of aging social infrastructure has become an issue, and it is important to switch from ex-post maintenance to preventive maintenance by monitoring these conditions.
Under such circumstances, monitoring using a strain gauge has been conventionally performed. This is a well-proven technique that is widely used to confirm the deformation of an object.

However, in the monitoring using the strain gauge, there is a problem that the measuring device is expensive and complicated wiring and power source are required.
Further, although monitoring using such a strain gauge is suitable for monitoring the state of a structure, it is not suitable for monitoring the state of a moving body such as a motorcycle.

JP, 2008-117423, A

Here, as a technique for monitoring the state of the moving body, the vehicle information acquisition method described in Patent Document 1 can be cited.
This method acquires accurate information of the vehicle only from the information from the sensor, and based on the acceleration information from the acceleration sensor provided in the vehicle, the state in which the vehicle is stationary and the state in which the vehicle is running. It is possible to determine whether or not the vehicle is being transported.

However, in this method, the state of the vehicle is determined based only on the information from the sensor, and therefore it may be difficult to identify the type of event that has occurred in the vehicle.
Further, since the vehicle status information is transmitted through the PHS line or the mobile phone line, a large capacity battery and an external power source are required.

Therefore, the present inventors have diligently studied and can monitor the states of both the moving body and the structure, and input the information for identifying the type of the event along with the state of the monitored object together with the information from the sensor. By creating time series information by using this time series information as learning data and performing machine learning to create a learned model, a program that can identify the type of event based on the information from the sensor is generated. Then, by installing this in the monitoring event analysis device, it became possible to analyze the event that occurred in the monitored object more appropriately.
In addition, in such a monitoring event analysis device, the event identification result can be transmitted using a power-saving wide-area wireless communication line (LPWA (Low Power Wide Area)), so that complicated wiring and an external power supply can be performed. It was unnecessary and enabled operation for several years.
Furthermore, in order to enable transmission of a large amount of data as needed, the monitoring event analysis device is equipped with a wireless switching unit so that transmission can be performed using a mobile phone line or the like.

The present invention has been made in view of the above circumstances, and it is possible to more appropriately identify the state of a monitored object such as a moving body or a structure without requiring complicated wiring or an external power source. It is an object of the present invention to provide a setting method of the monitoring event analysis device.

In order to achieve the above object, a method for setting a monitoring event analysis device according to the present invention is a monitoring event analysis device which is attached to a monitoring target object, identifies a type of an event occurring in the monitoring target object, and transmits an event identification result. The event identification processing unit, wherein the monitoring event analysis apparatus executes an event identification process based on a sensor for measuring the state of the monitored object and an event identification program to generate the event identification result. And an information collecting device that includes a wireless communication unit that transmits the event identification result and that includes a sensor that measures the state of the monitored object, and inputs sensor information by the information collecting device. At the same time, the event identification information corresponding to each type of event is input at the timing when the event occurs, time series information having the sensor information and the event identification information is created, and the time series information is obtained from the information collection device. Is transmitted to a monitoring information processing device via a communication line as learning data, the monitoring information processing device executes machine learning using the learning data, and the type of event is identified based on the sensor information. This is a method of generating an elephant identification program and installing the event identification program in the monitoring event analysis device.

ADVANTAGE OF THE INVENTION According to this invention, the setting method of the monitoring event analysis apparatus which enables more appropriately discriminating the state of the monitoring object, such as a mobile body and a structure, without requiring complicated wiring or an external power supply. Can be provided.

It is a block diagram which shows the structure of the monitoring system used in the setting method of the monitoring event analysis apparatus which concerns on embodiment of this invention. It is a block diagram which shows the structure of the information collection device used in the setting method of the monitoring event analysis apparatus which concerns on embodiment of this invention. It is a figure which shows the time series information used in the setting method of the monitoring event analysis apparatus which concerns on embodiment of this invention. It is a flow chart which shows a process of a setting method of a surveillance event analysis device concerning an embodiment of the present invention. It is a flowchart which shows the process of monitoring a monitoring target object using the monitoring event analysis apparatus in embodiment of this invention. It is a block diagram which shows the structure of the application example of the monitoring system used in the setting method of the monitoring event analysis apparatus which concerns on embodiment of this invention.

Hereinafter, an embodiment of the setting method of the monitoring event analysis device of the present invention will be described in detail. However, the present invention is not limited to the specific contents of the following embodiments.

First, the setting method of the monitoring event analysis apparatus of this embodiment will be described with reference to FIGS. FIG. 1 is a block diagram showing the configuration of a monitoring system used in the setting method of the monitoring event analysis apparatus of this embodiment. FIG. 2 is a block diagram showing the configuration of the information collection device used in the method for setting the monitoring event analysis device according to the present embodiment. FIG. 3 is a diagram showing time series information used in the setting method of the monitoring event analysis apparatus of this embodiment. FIG. 4 is a flowchart showing steps of the setting method of the monitoring event analysis device of this embodiment.

As shown in FIG. 1, the monitoring system used in the monitoring event analysis apparatus setting method of the present embodiment includes a monitoring event analysis apparatus 1, a communication line 2, a monitoring information processing apparatus 3, and an information collecting apparatus 4. .. In this monitoring system, a plurality of monitoring event analysis devices 1 can be preferably provided.

The monitoring event analysis device 1 is a so-called IoT device, is attached to a monitoring target, collects information of the target using various sensors, identifies the type of event occurring in the monitoring target, and outputs the event identification result. Is a device that transmits an acceleration signal, a temperature sensor 11, a data storage unit 12, an event identification program storage unit 13, an event identification processing unit 14, a wireless communication unit 15, a power supply unit 16, and a timer unit 17. ..

The object to be monitored may be an object that vibrates, and more specifically, it may be a moving object such as a motorcycle, a structure such as a mold, a door, a window, or the like.

The acceleration sensor 10 is a sensor that measures vibration information of a monitoring target and acquires acceleration, and is a single-axis or multi-axis accelerometer capable of detecting minute vibration.
As the multi-axis accelerometer, a 6-axis accelerometer that measures acceleration in the XYZ directions and angular velocity in the XYZ directions can be preferably used.
The temperature sensor 11 acquires the temperature of the monitored object.

It should be noted that the monitoring event analysis device 1 may include various other sensors. For example, a strain sensor, an angle sensor, an acoustic sensor, an ultrasonic sensor, a humidity sensor, a GPS sensor, a distance sensor, etc. are provided, sensor information can be acquired from these sensors, and the type of event can be identified based on the sensor information. You can also

The data storage unit 12 is a storage unit that stores the sensor information input from the acceleration sensor 10 and the temperature sensor 11, and can be configured by a storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive).
The data storage unit 12 also stores the event identification result generated by the event identification processing unit 14 and indicating the result of identifying the state (type of event) of the monitored object.

The event identification program storage unit 13 is a storage unit that stores an event identification program that identifies the type of event based on sensor information, and is configured by a storage device such as a HDD (Hard Disk Drive) or SSD (Solid State Drive). You can In addition, the event identification program storage unit 13 can store various other application programs and the like.

The event identification processing unit 14 executes the event identification processing using the sensor information based on the event identification program, and generates the event identification result.
For example, when the monitored object is a motorcycle, the event identification result may indicate right/left turn, acceleration/deceleration, fall, theft, engine start, engine drive time, and the like.

The wireless communication unit 15 is a communication interface having a chip in which an application for realizing communication with the communication line 2 is mounted and an antenna, and is connected to the communication line 2 as an LPWA (low power wide area) to connect data. Send and receive.
That is, since the event identification result generated by the event identification processing unit 14 can perform wireless communication with a small capacity, it is necessary to use the one using the LPWA network as the wireless communication unit 15 to transmit the event identification result. Is possible.

The event identification result can be generated as compressed data by the event identification processing unit 14, and is decompressed by the monitoring information processing device 3 or the terminal (not shown) of the user who connects to the monitoring information processing device 3 and uses the monitoring system. Can be used.

The data transmitted and received by the wireless communication unit 15 includes, in addition to the event identification result, a part of data regarding the state of the monitored object, abnormality information regarding the state of the monitored object, abnormality information generated in the monitoring event analysis device 1, and monitoring. It may be the control information or the like from the information processing device 3.
When the communication line 2 is, for example, SIGFOX, the amount of data at one time is up to 12 bytes at present. The wireless communication unit 15 transmits information that fits in such a data capacity to the monitoring information processing device 3 via the communication line 2.

The power supply unit 16 only needs to be able to supply power to the data storage unit 12, the event identification program storage unit 13, the event identification processing unit 14, the wireless communication unit 15, and the timer unit 17, and a power generation function for generating power from external energy. And a battery, a piezoelectric element that converts vibration of a structure into electric power, a heat conversion element that converts heat into electric power, a removable dry cell, a solar cell, and a dye-sensitized solar cell. The power supply unit 16 is preferably an independent power supply that does not require wiring or charging.

The timer unit 17 has a time counting function, can set a predetermined time and period arbitrarily, and outputs a control signal to the event identification processing unit 14 when the set value is reached.
Further, the monitoring event analysis device 1 may be provided with a voice output unit (not shown) so that a corresponding voice can be output based on a predetermined event identification result.

The communication line 2 includes a large-capacity data transmission line and a smaller-capacity data transmission line.
Cellular LPWA, mobile phone lines such as communication 3G and LTE, Bluetooth (registered trademark), Wi-Fi (registered trademark), EnOcean (registered trademark), ZigBee (registered trademark), etc. And a communication line of the short-range wireless communication system of.

On the other hand, for a small capacity data transmission line, LPWA (such as SIGFOX (registered trademark), LoRa (registered trademark), NB-IoT, NB-Fi Protocol, GreenOFDM, DASH7, RPMA, Wi-SUN, and LTE-MTC) is used. Low Power Wide Area) is included.
SIGFOX, which is an example of LPWA, has a long transmission distance of several tens of kilometers, a transmission speed of 100 bps (upstream) and an ultra-low speed, and data of 12 bytes (upstream) and 100 of Ethernet (registered trademark) data. It is less than a fraction of the size.

The monitoring information processing device 3 is an information collecting device such as a server or a computer that remotely monitors a monitored object, generates an event identification program using the learning data transmitted from the monitoring event analysis device 1, and Can be installed in the monitoring event analysis apparatus 1. In addition, a terminal (not shown) of a user who uses the monitoring system is connected to the monitoring information processing apparatus 3 through a wireless communication line, and based on the information output from the monitoring information processing apparatus 3 to the user, It is also possible to make it possible to grasp the state of the monitored object.

The monitoring information processing device 3 may include a storage unit 30, an event identification program generation unit 31, and a communication unit 32.
The storage unit 30 stores the event identification result transmitted from the monitoring event analysis apparatus 1, the learning data transmitted from the information collection apparatus 4, the event identification program generated by the event identification program generation unit 31, and the like. The storage unit is a storage unit such as a HDD (Hard Disk Drive) or SSD (Solid State Drive).

The event identification program generation unit 31 creates a learned model from the learning data and creates an event identification program.
Learning data and a predetermined machine learning algorithm are used for model learning. The learning data includes various information included in the time series information. Specifically, the feature amount extracted from the sensor information, the event identification information corresponding to the feature amount, and the time series displacement of these can be used. It can be said that these correspondences show the history of the state of the monitored object, and are suitable as learning data.

For example, the feature amount becomes an explanatory variable for learning the model, and the event identification information becomes an objective variable for learning the model. Therefore, by inputting the feature amount and outputting the event identification information, the event identification program generation unit 31 learns whether the objective variable can be explained by the explanatory variable, and learns and creates a model that can be quantitatively analyzed. can do.
The learned model generated by the event identification program generation unit 31 is reflected in the event identification program, installed in the monitoring event analysis device 1 by the monitoring information processing device 3, and stored in the event identification program storage unit 13.

The communication unit 32 is a communication interface including a chip in which an application for realizing communication with the communication line 2 is mounted and an antenna.

The information collecting device 4 measures the state of the monitored object and inputs the sensor information, and also inputs the event identification information corresponding to each type of the event at the timing when the event occurs to obtain the sensor information and the event identification information. The time series information that it has is created, and this is sent to the monitoring information processing device 3 via the communication line 2 as learning data. As will be described later, the monitoring event analysis device 1a used in the application example of the setting method of the monitoring event analysis device can also be used as the information collection device 4.

As shown in FIG. 2, the information collecting device 4 includes an acceleration sensor 40, a temperature sensor 41, an event identification information input unit 42, a time series information storage unit 43, a data processing unit 44, a wireless communication unit 45, a power supply unit 46, and A timer unit 47 is provided.
The acceleration sensor 40, the temperature sensor 41, the power supply unit 46, and the timer unit 47 can have the same configurations as the acceleration sensor 10, the temperature sensor 11, the power supply unit 16, and the timer unit 17 in the monitoring event analysis device 1.

The event identification information input unit 42 inputs the event identification information for identifying the state (type of event) of the monitored object and stores it in the time series information storage unit 43.
The event identification information input unit 42 can be configured by a switch, for example, and a person can press the switch at a timing when a predetermined event occurs to input the corresponding event identification information.
For example, if the monitored object is a motorcycle, attach a switch for right turn to the right handle, attach a switch for left turn to the left handle, press the switch for right turn when making a right turn, and make a switch when making a left turn. Each event identification information can be input by pressing the left turn switch.

The type of event is not particularly limited, but includes, for example, left/right turning of a motorcycle, acceleration/deceleration, falling, theft, engine start, engine running time, opening/closing of doors and windows, mold wear/abnormality of molds, etc. Can be mentioned.

The time-series information storage unit 43 stores time-series information as learning data. As shown in FIG. 3, the time-series information can include sensor information such as acceleration in the XYZ directions, angular velocity in the XYZ directions, temperature, and various event identification information at each time, for example.

As the event identification information, for example, when the monitored object is a motorcycle, information indicating a right turn or a left turn of the motorcycle can be used. Specifically, a flag is set by changing the corresponding event identification information from "0" to "1" when making a right turn, and the corresponding event identification information is changed from "0" to "1" when making a left turn. By setting the flag by changing to “1”, the event identification information corresponding to the sensor information can be recorded in the time series information.

Then, by using this event identification information as learning data in machine learning by the monitoring information processing device 3, it is possible to generate an event identification program that can identify the type of event based on the sensor information.
Further, by setting such event identification information variously for each monitoring target object, according to the setting method of the monitoring event analysis device of the present embodiment, for various monitoring target objects, the type of event based on the sensor information. It is possible to obtain a monitoring event analysis device that can identify

In addition, depending on the types of the monitored object, the sensor information, and the event identification information, it may be difficult to simultaneously acquire the sensor information and the event identification information and create the time-series information.
In consideration of such a case, it is also preferable that the sensor time-series information having the sensor information and the event time-series information having the event identification information are separately created and transmitted to the monitoring information processing device 3. ..
For example, an event identification information input device (not shown) is provided separately from the information collection device 4, and an event identification information input unit, a time-series information storage unit, a data processing unit, and a wireless device similar to those in the information collection device 4 are provided therein. It is also preferable to provide a communication unit, a timer unit, and a power supply unit, and create event time series information having event identification information for each time, separately from the sensor time series information.

At this time, the data processing unit 44 of the information collection device 4 transmits the sensor time series information from the wireless communication unit 45 to the monitoring information processing device 3, and the data processing unit of the event identification information input device receives the event time series information from the wireless communication unit. Is transmitted to the monitoring information processing device 3.
Then, the storage unit 30 of the monitoring information processing device 3 stores the transmitted sensor time-series information and event time-series information, and the event identification program generation unit 31 matches them at each time to create learning data. It is also preferable that the machine learning is performed to generate the event identification program.
The same applies when the monitoring event analysis device 1a is used as the information collection device 4.

The data processing unit 44 transmits the time series information stored in the time series information storage unit 43 as learning data from the wireless communication unit 45 to the monitoring information processing device 3.
The wireless communication unit 45 is a communication interface having a chip on which an application for realizing communication with the communication line 2 is mounted and an antenna, and is connected to the mobile phone line or the communication line 2 of the data short-range wireless communication system. Send and receive data. That is, the wireless communication unit 45 can transmit and receive a large amount of data by connecting to such a line or network.

Note that the wireless communication unit 45 is omitted from the information collecting device 4, the monitoring information processing device 3 is connected to the information collecting device 4 by wire, and learning data is transmitted from the information collecting device 4 to the monitoring information processing device 3. You may

The setting method of the monitoring event analysis apparatus of this embodiment has the following steps, as shown in FIG.
First, the information collecting device 4 is attached to the monitored object (step 10).
Next, the acceleration sensor 40 and the temperature sensor 41 in the information collecting device 4 measure the acceleration, the angular velocity, and the temperature, and these sensor information are input and stored in the time series information storage unit 43.
Further, the event identification information input unit 42 of the information collection device 4 inputs the event identification information corresponding to each event type at the timing when a predetermined event occurs, and stores the event identification information in the time series information storage unit 43 ( Step 11).
As a result, the time-series information storage unit 43 creates the time-series information as shown in FIG. 3 as learning data (step 12).

Next, the data processing unit 44 of the information collecting device 4 transmits the learning data from the wireless communication unit 45 to the monitoring information processing device 3 (step 13).
The monitoring information processing device 3 stores the learning data in the storage unit 30, machine-learns the event identification program generating unit 31 to create a learned model from the learning data (step 14), and based on the sensor information. An event identification program for identifying the type of event is generated (step 15).

Next, the event identification program generated by the event identification program generation unit 31 in the monitoring information processing device 3 is converted into an OS (operating system) independent language and stored in the storage unit 30 (step 16).
Then, the converted event identification program is installed in the monitoring event analysis device 1 and stored in the event identification program storage unit 13 (step 17). At this time, the event identification program can be installed by connecting the monitoring information processing device 3 to the monitoring event analysis device 1 by wire.

As described above, the event identification program is converted into an OS (operating system) independent language, then installed in the monitoring event analysis apparatus 1, and the event identification processing is performed based on the event identification program. It is possible to significantly reduce the amount of power consumed in the monitoring event analysis apparatus 1 as compared with the case of executing.

Next, a process of monitoring a monitoring target using the monitoring event analysis device 1 obtained in this way will be described with reference to FIG.
First, the monitoring event analysis device 1 is attached to the monitored object (step 20).
Then, the monitoring event analysis device 1 collects sensor information such as acceleration, angular velocity, and temperature, and stores it in the data storage unit 12 (step 21).

Next, the event identification processing unit 14 of the monitoring event analysis apparatus 1 executes the event identification program of the event identification program storage unit 13 to determine the corresponding event based on the sensor information, generate the event identification result, and generate the data. It is stored in the storage unit 12 (step 22). The event identification result can be generated, for example, as information indicating that the event A has occurred once, the event B has occurred 12 times, the event C has occurred 3 times, and the event D has occurred 0 times.
Execution of the event identification program can be executed at a fixed timing set by the timer unit 17 or at any time based on instruction information from an event identification instruction unit (not shown) in the monitoring information processing apparatus 3. it can.

Then, the wireless communication unit 15 in the monitoring event analysis apparatus 1 uses the LPWA network to transmit the event identification result (step 23). The timing of transmitting the event identification result can be set to be executable immediately after the execution of the event identification program or at any time based on the instruction information from the monitoring information processing apparatus 3.
The monitoring information processing device 3 receives the event identification result (step 24) and displays it on the display unit (not shown) (step 25).

According to the setting method of the monitoring event analysis apparatus of this embodiment, the state of the monitoring target such as the moving body or the structure is more appropriately identified without the need for complicated wiring or an external power source. It becomes possible.

Next, an application example of the monitoring event analysis device according to the present embodiment and an application example of the setting method of the monitoring event analysis device using the same will be described.
The monitoring event analysis apparatus 1a in this application example is different from the monitoring event analysis apparatus 1 in that it can perform large-capacity wireless communication, and the other points are the same as the monitoring event analysis apparatus 1.

As shown in FIG. 6, the monitoring event analysis device 1a includes an acceleration sensor 10a, a temperature sensor 11a, a data storage unit 12a, an event identification program storage unit 13a, an event identification processing unit 14a, a first wireless communication 15a, and a power supply unit. 16a, a timer unit 17a, a second wireless communication unit 18a, and a wireless switching unit 19a.

The first wireless communication unit 15a may be the same as the wireless communication unit 15 in the monitoring event analysis device 1. Further, the configuration other than the second wireless communication unit 18a and the wireless switching unit 19a can be the same as the configuration with the same name in the monitoring event analysis device 1.

The second wireless communication unit 18a is a communication interface having a chip in which an application for realizing communication with the communication line 2 is mounted and an antenna, and is a mobile phone line or a communication line 2 of a data short-range wireless communication system. Connect to to send and receive data. The second wireless communication unit 18a can transmit and receive a large amount of data by connecting to such a line or network. As the large amount of data, time-series information or an event identification program, which is learning data, can be used.

The wireless switching unit 18a switches the wireless communication unit to be used to the first wireless communication unit 16a or the second wireless communication unit 18a. The wireless switching unit 18a can switch the wireless communication unit that transmits the event identification result to the first wireless communication unit 16a or the second wireless communication unit 18a based on the event identification result. In addition, the wireless switching unit 18a determines which wireless communication unit to use based on the switching instruction information transmitted from the communication switching instruction unit (not shown) in the monitoring information processing device 3 as the first wireless communication unit 16a or the first wireless communication unit 16a. It is possible to switch to the second wireless communication unit 18a.

The wireless switching unit 18a switches the first wireless communication unit 16a to the second wireless communication unit 18a, for example, when the event identification result indicates an abnormality of the monitored object, so that the event identification processing unit 14a causes the event identification It is possible to transmit the corresponding time series information together with the result from the second wireless communication unit 18a to the monitoring information processing device 3. Further, in the monitoring information processing device 3, it is possible to create a learned model by using this time series information as learning data and reflect it in the event identification program.

In addition, the wireless switching unit 18a switches the first wireless communication unit 16a to the second wireless communication unit 18a when the period for collecting the sensor information from the state of the monitored object reaches a predetermined value to identify the event. The processing unit 14a can also cause the second wireless communication unit 18a to transmit the time-series information to the monitoring information processing device 3.
Furthermore, when the transmission of the large amount of data is completed, the wireless switching unit 18a switches the second wireless communication unit 18a to the first wireless communication unit 16a. With this, when only the event identification result is transmitted, the first wireless transmission unit 16a can be used, and the power consumed in the monitoring event analysis device 1a can be reduced.

It should be noted that the wireless switching unit 18a may switch the second wireless communication unit 18a to the first wireless communication unit 16a based on the event identification result when only the event identification result is transmitted.

Furthermore, it is also preferable that the monitoring event analysis device 1a includes the event identification information input unit 42 and the time-series information storage unit 43 of the information collection device 4.
With such a configuration of the monitoring event analysis device 1a, the monitoring event analysis device 1a can be used as the information collecting device 4 in the above-described monitoring event analysis device setting method, and can also be used for monitoring the monitored object.

According to such a monitoring event analyzing apparatus 1a, for example, when the monitoring event analyzing apparatus 1a is attached to a monitored object and then it is found that the identification rate of the event type is not so good, the monitoring event analyzing apparatus 1a It is possible to transmit the learning data from the monitoring event analysis device 1a to the monitoring information processing device 3 in a state where is attached to the monitoring target. Then, the event identification program generated in the monitoring information processing device 3 is transmitted from the communication unit 32, received by the second wireless communication unit 18a of the monitoring event analysis device 1a, and stored in the event identification program storage unit 13a. You can
Therefore, the event identification program can be updated while the monitoring event analysis device 1a is attached to the monitoring target, and a more excellent monitoring event analysis device 1a can be provided.

As described above, according to the application example of the setting method of the monitoring event analysis apparatus according to the present embodiment, the state of the monitored object can be more appropriately identified without the need for complicated wiring or external power supply. It is possible, and the normal event identification result can be transmitted using the LPWA network with a small capacity, and only when the time series information and the event identification program need to be transmitted and received, a large capacity line can be transmitted. Can be switched to use.

It is needless to say that the present invention is not limited to the above embodiment and various modifications can be made within the scope of the present invention.
For example, the present invention is not limited to the configuration shown in FIG. 1 and the like as long as the monitoring system has a function capable of executing the above-described monitoring method. Further, each of the above steps can be executed by hardware or software. Furthermore, one functional block may be configured by hardware alone, software alone, or a combination thereof. Further, it is possible to make appropriate changes such as a configuration in which the respective configurations in FIG. 1 and the like are further combined.

INDUSTRIAL APPLICABILITY The present invention can be preferably used to more appropriately identify the state of a monitoring target such as a moving body or a structure without requiring complicated wiring or an external power source. ..

1, 1a Monitored event analysis device 10, 10a Acceleration sensor 11, 11a Temperature sensor 12, 12a Data storage unit 13, 13a Event identification program storage unit 14, 14a Event identification processing unit 15 Wireless communication unit 15a First wireless communication unit 16 , 16a Power supply unit 17, 17a Timer unit 18a Second wireless communication unit 19a Wireless switching unit 2 Communication line 3 Monitoring information processing device 30 Storage unit 31 Communication unit 32 Learned model generation unit 33 Wireless switching instruction unit 4 Information collecting device 40 Acceleration sensor 41 Temperature sensor 42 Event identification information input unit 43 Time series information storage unit 44 Data processing unit 45 Wireless communication unit 46 Power supply unit 47 Timer unit

Claims (6)

A method of setting a monitoring event analysis device, which is attached to a monitoring object, identifies the type of event that has occurred in the monitoring object, and transmits an event identification result,
The monitoring event analysis device, a sensor that measures the state of the monitored object, an event identification processing unit that executes event identification processing based on an event identification program to generate the event identification result, and the event identification result Equipped with a wireless communication unit for transmitting
An information collecting device having a sensor for measuring the state of the monitoring target is attached to the monitoring target,
The sensor information is input by the information collecting device, and the event identification information corresponding to each event type is input at the timing when the event occurs to create time series information including the sensor information and the event identification information. ,
The time-series information is transmitted from the information collecting device to the monitoring information processing device via a communication line as learning data,
Machine learning is performed by the monitoring information processing apparatus using the learning data, the event identification program for identifying the type of event is generated based on the sensor information, and the event identification program is used as the monitoring event analysis apparatus. A method for setting up a monitoring event analysis device, characterized by being installed in. 2. The monitoring information processing apparatus converts the event identification program into an OS (operating system) independent language, and the converted event identification program is installed in the monitoring event analysis apparatus. For setting up the monitoring event analysis device of. The monitoring event analysis according to claim 1 or 2, wherein the wireless communication unit transmits the event identification result using a power-saving wide-area wireless communication line (LPWA (Low Power Wide Area)). How to set up the device. The sensor is an acceleration sensor that acquires acceleration and angular velocity of the monitored object,
The monitoring event analysis device setting method according to any one of claims 1 to 3, wherein the event identification information includes information indicating a right turn or a left turn of the motorcycle. The monitoring event analysis device switches a second wireless communication unit that uses a mobile phone line or a short-distance wireless communication communication line, and switches the wireless communication unit to be used to the wireless communication unit or the second wireless communication unit. The method for setting a monitoring event analysis device according to claim 1, further comprising a wireless switching unit. Using the monitoring event analysis device as the information collecting device, the monitoring event analysis device is attached to the monitored object,
Sensor information is input by the monitoring event analysis device, and event identification information corresponding to each event type is input at the timing when the event occurs, and time series information having the sensor information and the event identification information is created. Then
The monitoring event analysis apparatus setting method according to claim 5, wherein the monitoring event analysis apparatus transmits the time-series information as learning data to a monitoring information processing apparatus via a communication line.

Images