KR20220083181A - Apparatus and method for selecting collected data for smart farm dataset validation - Google Patents

Abstract

The present invention relates to an apparatus and method for selecting collected data for validating smart farm datasets.
The collected data selection device for smart farm dataset validation according to an embodiment of the present invention includes an input unit for receiving data generated from smart farm equipment, a memory and a program in which a program for selecting data based on a ruleset is stored. and a processor that executes, wherein the processor selects data by applying a rule set that is a set of rules for data selection.

Description

Apparatus and method for selecting collected data for smart farm data set validation

The present invention relates to an apparatus and method for selecting collected data for validating smart farm datasets.

According to the prior art, there is a problem that the proportion of unnecessary data is very high when collecting data necessary for building big data.

As the amount of unnecessary data increases, the waste of resources for data processing increases, and there is a problem in that reliability is not secured because the results are distorted in data analysis.

The present invention has been proposed to solve the above problems, and the purpose of the present invention is to provide an apparatus and method for selecting collected data that selects data necessary for building big data through validation of data generated and used in a smart farm. have.

An apparatus for selecting collected data for validation of a smart farm dataset according to the present invention includes an input unit for receiving data generated from smart farm equipment, a memory in which a program for selecting data based on a ruleset is stored, and a processor for executing the program Including, the processor is characterized in that the data is selected by applying a rule set, which is a set of rules for data selection.

According to the present invention, by verifying the validity of the data collected for smart farm big data construction and storing and maintaining it through selection, the big data construction cost is reduced and unnecessary data transmission is reduced to analyze the data collection node and data It reduces the load on the system and has the effect of reducing the network cost required for data transmission.

According to the present invention, when building smart farm big data, data is selected through ruleset definition according to the pre-processing selection method in all subjects of the data collection chain leading to the smart greenhouse (including livestock and field farms), farms, and cloud, and the purpose of data collection It is possible to collect high-quality data in accordance with the

In addition, when applied to the lower end of the data collection chain, such as a smart greenhouse, it is possible to effectively reduce the cost of data collection by avoiding unnecessary waste of system resources at the upper end in advance.

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

1 shows the configuration of a smart farm in an embodiment of the present invention.
Figure 2 shows a smart farm data collection procedure according to an embodiment of the present invention.
3 shows a smart farm data selection function block according to an embodiment of the present invention.
4 shows an apparatus for selecting collected data for smart farm dataset validation according to an embodiment of the present invention.

The above and other objects, advantages and features of the present invention, and a method for achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings.

However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only the following examples are provided to those of ordinary skill in the art to which the present invention pertains. It is only provided to easily inform the composition and effect, and the scope of the present invention is defined by the description of the claims.

On the other hand, the terms used herein are for the purpose of describing the embodiments and are not intended to limit the present invention. In this specification, the singular also includes the plural unless specifically stated otherwise in the phrase. As used herein, "comprises" and/or "comprising" means that a referenced element, step, operation and/or element is the presence of one or more other elements, steps, operations and/or elements. or added.

Hereinafter, in order to facilitate the understanding of those skilled in the art, a background in which the present invention is proposed will be first described, and embodiments of the present invention will be described.

Recently, as smart farms spread, the importance of data generated during smart farm operation is gradually increasing.

In particular, with the development of machine learning technologies such as deep learning, the possibility that primary industrial agricultural technology, which was only dependent on existing human resources, can have secondary industrial characteristics, has increased, and the importance of smart farm data is rising. .

As technologies such as artificial intelligence are used in a variety of ways, the role of data grows larger, because data is one of the key components in the operation of artificial intelligence.

It is easy to predict that the importance of smart farm data will gradually grow in the future because the more and better data we have, the better the AI will produce quality results.

Since securing more high-quality data ultimately leads to future competitiveness, various types of data collection are being actively conducted in various fields at home and abroad.

However, in many cases, redundant or unnecessary data is collected.

In general, when collecting data necessary for building big data, the proportion of unnecessary data is very high enough to say that more than 80% of it is garbage data.

As the amount of unnecessary data that is not the user's desired data increases, the waste of resources for data processing becomes more severe, and there are many obstacles in avoiding distorted results and providing reliable and accurate results in data analysis.

Accordingly, there is a need for an efficient data collection method, which can be divided into a pre-processing screening method and a post-processing screening method depending on the time point involved in data collection.

The post-processing screening method is a method of efficiently storing the collected data by compressing the volume without being involved in data collection.

That is, the data collection process is simple as a method of collecting all data once and selecting them later, but data validation must be performed separately.

It does not cost a lot to collect data right away, but for data analysis, data selection and verification process is required.

In general, it can be applied when collecting data through simple devices and performing data validation and analysis through high-performance computing resources such as the cloud.

The pre-processing screening method is a method of selecting and storing only the data that is worth saving by selecting the data to be collected from the data collection process.

Since data needs to be selected during the data collection process, a data collection device with higher performance than the post-processing screening method is required, and logic implementation for what purpose the data will be collected for is also required.

However, the data construction method through efficient pre-processing selection enables the efficient use of resources, and since the data itself is data collected for the purpose, high-quality results can be expected in data analysis, so the value of the data can be evaluated more highly. .

Since the pre-processing and post-processing screening methods each have their pros and cons, it is appropriate to select and apply them according to the conditions and purpose for data collection.

The present invention has been proposed to solve the above-described problems, and according to an embodiment of the present invention, an efficient data collection apparatus and method for smart farm big data construction are provided, and unnecessary data generated in the collection process is removed, It is possible to ultimately obtain high-quality data analysis results by minimizing the cost burden in the subsequent process and enhancing the value of data according to the purpose.

Specifically, according to an embodiment of the present invention, a data collection device and method related to preprocessing selection when constructing smart farm big data are proposed.

Data in the field of smart farms have a very wide range of data collection, such as smart greenhouses, smart livestock houses, and smart orchards, and as the types of crops and livestock are diverse, the types of data also have various characteristics.

Therefore, the need for data collection according to data selection is continuously emerging, and it is easily predicted that the equipment currently applied to the smart farm will also be upgraded continuously with the continuous development of ICT technology. It is expected that the method can be used effectively.

Smart farm is a cutting-edge agricultural technology that seeks to create high added value by improving productivity and efficiency by converging ICT technologies such as IoT with existing traditional agriculture, and various IT devices are used.

Smart farms can be classified into smart greenhouses for growing crops, smart barns for raising livestock, and smart orchards for growing fruit trees, depending on the field of application.

That is, smart farms can be classified according to the growing object, and subdivided again according to the breed of the growing object.

As such, various types of facilities and devices may be provided in a smart farm depending on the type and method of farming, and smart farm data is generated in very different types in each case.

1 shows the configuration of a smart farm in an embodiment of the present invention.

As shown in Figure 1, each smart greenhouse has a sensor module (temperature sensor module, humidity sensor module, illuminance sensor module, etc.), driver module (relay, motor, etc.), a sensor node 114 and a driver node 117 and , a control node 118 that receives environmental and equipment status data of the greenhouse from them and transmits control commands for them is provided.

Sensor types can be provided in various ways depending on the data to be measured, such as Temperature, Humidity, and Illuminance. It may be provided in various ways.

A plurality of smart greenhouses may be provided in a farmhouse, and through a control node provided for each smart greenhouse, it is connected to the operation node 120 in charge of the overall operation of the farmhouse.

Each control node 118 transmits a signal to control the driver node 117 , and also transmits data transmitted by the sensor node 114 and the driver node 117 to the operation node 120 .

Since the operating node 120 is used like an operating system that overall observes and integrally controls the state of each greenhouse operated by the farmhouse by integrating and managing a plurality of greenhouses, the smart farm data generated by the farmhouse is the operating node 120 ) can be collected and managed in

The operation node 120 of each farmhouse may be connected to the management system 130 of the cloud. In this case, the management system 130 identifies and analyzes the operation status of a plurality of remote farmhouses to find the optimal operating conditions and again each farmhouse It is possible to feedback through the operating node 120 of

Accordingly, data collected from each farm is collected and managed in the cloud management system 120 .

According to this method, the cloud can provide various smart farm services.

Examples of such smart farm services include a smart farm data analysis service that derives an optimal growth model by analyzing growth and environmental information generated in a smart greenhouse, and a remote analysis of the growth status of crops in the smart greenhouse by sending an image. This includes remote diagnostic services that identify conditions or pests.

The above-described service can be performed using its own data in the operation node 120 in the farmhouse, but a difference occurs from the analysis result based on a large amount of data processed in the cloud management system 130 .

The data that can be generated and collected in a smart farm will vary depending on the type of smart farm, such as greenhouses, livestock houses, and open fields, but can be broadly classified into four fields: facility, environment, growth, and management based on smart greenhouses.

Facility data is data for basic information of smart farm facilities such as smart greenhouse, and includes farmhouse information, facility information, and person in charge information.

Since these facility data are mainly used for the purpose of smart farm management, the data once created has the characteristic that it is not easily changed.

Environmental data is information generated from smart farm equipment such as sensors and actuators inside and outside the facility, and has a characteristic that is automatically measured according to a set period.

Growth data is information that is measured and generated according to the growth of crops, and has very diverse characteristics depending on the type of crop.

For example, even for facility horticultural crops grown in the same smart greenhouse, the types of growth data are very diverse depending on the types of crops such as tomatoes, paprika, and strawberries.

There are attempts to automate the collection of growth data through the introduction of artificial intelligence technology, but so far, growth data is mainly measured manually by skilled personnel.

Management data is information related to costs and profits required for smart farm management, and includes production costs, harvest information, and sales information.

Since management information reveals the profitability of farms, it is usually not disclosed well, and it is often estimated through market information such as auctions.

As described above, there is a wide variety of data that can be created and managed while operating a smart farm.

Among these, facility data and management data are usually entered manually rather than using a separate data collection device because the cycle of data generation is unclear. It is more preferable to apply

In the case of growth data, some are attempting automatic data collection using artificial intelligence technology, but it is still common to manually check and input data. It can be said that the post-processing screening method is more suitable.

However, in the case of environmental data, the spread of smart greenhouses has become common, and due to this, sensors that measure temperature, humidity, CO2, solar radiation, wind direction, wind speed, rain, photon, soil moisture content, soil moisture tension, EC, pH, geo temperature, etc. A wide variety of smart farm data is continuously generated because various smart farm equipment including skylights, side windows, thermal insulation covers, shading curtains, ventilators, floating fans, irrigation motors, irrigation valves, and actuators such as air conditioners are being produced, distributed and utilized. have.

According to an embodiment of the present invention, the selection of data generated by smart farm equipment will be described as an example.

The smart farm data collection procedure for big data utilization is carried out through data collection, data selection, data delivery and data disposal.

When data is first created, whether or not to process the data sorting is determined according to the data sorting method.

In the case of the post-processing selection method, data is transferred to the next step without data selection, and in the case of the pre-processing selection method, a data selection process is performed.

The data selection process is made through various conditions such as the type of data and the purpose of data use.

Data that is not suitable for the type of data, is duplicate data, or does not meet the purpose of data collection or utilization is discarded through the data disposal process. It is transferred to the next step through the data transfer process for

Figure 2 shows a smart farm data collection procedure according to an embodiment of the present invention.

Smart farms can be divided into various fields according to application fields such as greenhouses, livestock pens, and open fields, and here again, they can be divided into various types according to growth targets.

Although equipment may be applied differently depending on each field and target of growth, the generation of smart farm data is performed in smart farm equipment including sensors and actuators.

In other words, regardless of what type of equipment the smart farm equipment is, it can be said that the smart farm equipment has the conditions to generate smart farm data.

Smart farm equipment, which are usually installed in greenhouses, livestock pens, and fields, etc. at the time of initial installation, sets standards for data generation, such as what data, at what period, in what unit, and by when, depending on the type of equipment.

After that, the smart farm equipment that has been set up will periodically generate data from the moment it is operated.

In generating smart farm data, the sensor and actuator module continuously measure the corresponding type of environment and state data (S201).

The measured value is transmitted to the sensor and driver node to which the corresponding sensor and driver module are connected.

In general, a plurality of sensor modules or actuator modules are connected to one sensor node or actuator node, which differs depending on the configuration of the smart farm facility.

The analog value measured and sent by the sensor and driver module is digitized by the sensor and driver node to generate digital data that is the target of data selection (S202).

The data generated by the sensor and driver nodes go through a data selection process according to the preprocessing selection method (S203).

Since the data screening process varies depending on the purpose of the data, the purpose of data utilization must be clear above all else to select valid data.

That is, there must be a definition of what data is required, the effective range of the data must also be specified, and the creation time of the data must also be specified.

This becomes an essential element for comparison with other types of data for future data analysis.

For example, in order to obtain data for deriving the optimal growth algorithm for crops, there is not only crop growth data, but also environmental data of facilities maintained and managed during the growing period of the crop.

In order to derive the optimal growth algorithm, crop growth data and environmental data of facilities are subdivided again, and after each causal relationship is analyzed for each subdivided item, an environment to which each item is applied differently is artificially constructed and Afterwards, the optimal algorithm is derived by comparing it with the growth results of crops.

In other words. As a result, an optimal growth algorithm is derived based on the results of analysis of the dynamics between crop growth data and facility environmental data.

Data determined to be invalid among the data generated by the sensor and driver nodes through the data selection process are discarded, and the data determined to be valid data are transferred to the connected operation node for use in data analysis (S204).

If necessary, history information on the data selection process, such as the type of data, the amount of generated data and discarded data, etc. may be separately stored for future performance measurement of the device.

Data selected in units such as smart greenhouses, smart livestock houses, and smart farms are transmitted to the operation node, which is an integrated data storage place within the farm, and the operation node collects and collects data received from each control node (S205).

The collected data is stored (S207) and used for data analysis as data within the farmhouse (S208), or transmitted to the management system of the cloud through the Internet (S209).

Whether or not the data to be transmitted to the management system is data to be transmitted to the management system is determined through the data selection process according to the purpose of the service provided by the management system.

The management system collects the data received from the operation nodes of each farmhouse (S210), and through the selection process of the collected data (S211), extracts and utilizes valid data for data analysis (S212).

3 shows a smart farm data selection function block according to an embodiment of the present invention.

The smart farm data sorting apparatus according to an embodiment of the present invention is composed of software that operates based on a ruleset, and is mounted on a control node, an operating node, a management system, etc. that perform a data sorting process.

The smart farm data selection function block according to an embodiment of the present invention includes an input interface 301 for data input and an output interface 307 for data output.

The ruleset management unit 304 performs an operation of adding, searching, modifying, and deleting a rule set that defines a criterion for determination for data selection for each input data and a processing rule for data that meets the corresponding condition. do.

A ruleset is a set of rules for selecting data, and which data is selected depends on which ruleset is applied.

In the end, data selection criteria vary depending on how the operator defines the ruleset.

Examples of rules for screening data include whether data is received at a reasonable time, whether data is received within an expected period registered in the data sorting device, whether it is received slower or faster than a set sensing period, whether data is valid for which data is set There are rules such as whether the data is within the range, whether the data is out of the measurement error range within the reference time, whether it is valid compared to other related data, and whether the measured values of the redundant sensors are different.

The user interface 303 is in charge of an interface for performing an operation on the ruleset.

The data caching unit 302 serves as a temporary buffer for input data.

This makes it possible to operate on the input data state.

For example, the average value of the previous 10 data of the temperature sensor #1 can be applied to the ruleset.

The ruleset application unit 305 performs a process of selecting valid data by applying a defined rule set to input data.

The history management unit 306 performs history storage for the input data and the selected data.

For example, the number of input data and discarded data for the last one day of a specific sensor module is stored.

When the data discard rate is ideally high or low, the operator checks the status of the device.

4 shows an apparatus for selecting collected data for smart farm dataset validation according to an embodiment of the present invention.

An apparatus for selecting collected data for validation of smart farm datasets according to an embodiment of the present invention includes an input unit 410 for receiving data generated from smart farm equipment, and a memory in which a program for selecting data based on a ruleset is stored 420 and a processor 430 for executing a program, wherein the processor 430 selects data by applying a rule set that is a set of rules for data selection.

The processor 430 performs an operation on a rule set that defines a condition used as a criterion for data selection and a processing rule for data that meets the corresponding condition.

The processor 430 determines whether data is received at an appropriate time, whether data is received within an expected period registered in the data sorting device, whether it is received slower or faster than a set sensing period, whether data is within a set validity range , defines a rule including at least one of whether the data is out of the measurement error range within the reference time, whether it is valid compared to other related data, and whether the measured values of the redundant sensors are different.

The processor 430 performs history storage for the input data and the selected data.

When the data discard rate is out of a preset range, the processor 430 transmits a device status check notification.

Meanwhile, the collection data selection method for smart farm dataset validation according to an embodiment of the present invention may be implemented in a computer system or recorded in a recording medium. A computer system may include at least one processor, memory, a user input device, a data communication bus, a user output device, and storage. Each of the above-described components performs data communication through a data communication bus.

The computer system may further include a network interface coupled to the network. The processor may be a central processing unit (CPU) or a semiconductor device that processes instructions stored in a memory and/or storage.

The memory and storage may include various types of volatile or non-volatile storage media. For example, memory may include ROM and RAM.

Therefore, the collection data selection method for smart farm dataset validation according to an embodiment of the present invention may be implemented as a computer-executable method. When the collection data selection method for smart farm dataset validation according to an embodiment of the present invention is performed on a computer device, computer readable instructions determine the collection data selection method for smart farm dataset validation according to the present invention can be done

On the other hand, the method for selecting collected data for smart farm dataset validation according to the present invention described above may be implemented as a computer readable code on a computer readable recording medium. The computer-readable recording medium includes any type of recording medium in which data that can be read by a computer system is stored. For example, there may be a read only memory (ROM), a random access memory (RAM), a magnetic tape, a magnetic disk, a flash memory, an optical data storage device, and the like. In addition, the computer-readable recording medium may be distributed in computer systems connected through a computer communication network, and stored and executed as readable codes in a distributed manner.

Claims (1)

an input unit for receiving data generated from smart farm equipment;
a memory in which a program for selecting data based on a ruleset is stored; and
a processor for executing the program;
The processor performs selection on the data by applying a rule set that is a set of rules for data selection
Collected data screening device for smart farm dataset validation.

Images