WO2021096451A1

WO2021096451A1 - An intelligent control system for greenhouse air conditioning

Info

Publication number: WO2021096451A1
Application number: PCT/TR2019/051146
Authority: WO
Inventors: Abdurrahman Ozgur POLAT; Ercan AVSAR
Original assignee: Cukurova Universitesi Rektorlugu
Priority date: 2019-11-14
Filing date: 2019-12-20
Publication date: 2021-05-20

Abstract

Thanks to the software it contains, the present invention related to a system that can control windows on greenhouses very close to one another based on the information it receives from the sensors and meteorology it contains, thus providing the necessary ventilation in all greenhouses.

Description

AN INTELLIGENT CONTROL SYSTEM FOR GREENHOUSE AIR CONDITIONING

TECHNICAL FIELD Thanks to the software it contains, the present invention related to a system that can control windows on greenhouses very close to one another based on the information it receives from the sensors and meteorology it contains, thus providing the necessary ventilation in all greenhouses. PRIOR ART

In some regions in our country (example: Mersin-Limonlu) greenhouses are built very closely together, creating ventilation problems for these greenhouses. In other words, even if the ventilation windows of the greenhouse are open, ventilation cannot be adequately provided that the windows of other greenhouses near the greenhouse are closed. The acquaintances of the neighbouring greenhouse owner request from the other and turn on the ventilation of the neighbouring greenhouse. However, it is not possible for the next or the end greenhouse to be aware of the situation to intervene Therefore, this problem has not been fully resolved. BRIEF DESCRIPTION OF THE INVENTION

The present invention allows for coordinated ventilation between greenhouses, given the problem mentioned in the background art, if the distance between greenhouses greater than one and a half meters is close to two meters. This is performed by checking the windows which are connected to the opening and closing mechanisms software, after evaluation of the data it receives via the sensors placed in the greenhouses and the meteorology. This makes the air inside the greenhouse ideal.

LIST OF FIGURES

Figure 1. Representation of Wind Directions by Scenario 1 Figure 2. Representation of Windows Opened by Scenario 1

Figure 3. Representation of Wind Directions by Scenario 2. Figure 4. Representation of Windows Opened by Scenario 2.

DETAILED DESCRIPTION OF THE INVENTION

If the distance between greenhouses of more than one and a half meters is close to two meters, the present invention uses the following data during decision-making phase in order for the ventilation to be performed in coordination among the greenhouses.

• The location, direction and forms of the greenhouses included in the system (used to direct the air through the geographical location and shape)

• Product being grown in each greenhouse

• Instantaneous temperature and humidity in a greenhouse

• Instant air flow rate from greenhouse windows

• Instantaneous direction and severity of the outdoor wind

• Meteorology data

From this data, greenhouse locations and product information are considered to be fixed data, temperature humidity and wind information as variable data. Fixed data must be entered manually (manually) by the user in the system. This means that no sensor is used to obtain this data. However, various sensors must be used to collect the variable data. Temperature and humidity sensors are used to make in-greenhouse measurements. An anemometer is used for the airflow rate in the windows, and a wind station (anemometer and heading sensor) is used to conduct measurements relating to the open-air wind. At this point, the Internet connection module of the system is activated. Via an internet connection, the data of the meteorological directorate to be received over the Internet and the data measured for the open-air are verified. In addition, a wireless network is installed between the greenhouses to collect variable data from different greenhouses on a central administrator computer. The long-distance wide area network (LoraWan - Long Range Wide Area Network) technology will be used to build this network structure and establish an Internet connection. This technology is ideal for the system when it comes to the present invention due to low energy consumption and wide coverage.

The system is subject to the following components:

• Temperature and humidity sensor

• Wind speed (anemometer) and direction sensor • LoraWan node points and LoraWan gateway

• Software

• Cloud server Technical Specifications:

• Temperature and Humidity Sensor

• Wind Speed and Direction Sensor

· LoraWan nodes and LoraWan gateway

• Software

It is used to provide real-time provisioning of system data, the manual response when needed, and to draw historical data from the cloud server to generate statistical information and to make automated response decisions.

Method:

With the system in the present invention, it is calculated how long and how much the window of which greenhouses must remain open to ensure that optimal ventilation in the greenhouses is coordinated. Based on the rate of change in the instantaneous temperature and humidity information collected from the greenhouses, it is determined approximately how long the corresponding greenhouse should be aerated. The system has also been developed to account for the different crops produced in greenhouses and uses different thresholds for temperature and humidity according to the product in the greenhouse. This means that even if the values in the greenhouses are close together, it is possible to decide that they are vented according to the product in the greenhouses. (For example, 60% humidity in a greenhouse containing an X product does not require ventilation, while 55% moisture in the neighbouring greenhouse with Y product may cause aeration.) After the windows remain open for the calculated time in the cases mentioned above, the data in the greenhouse is measured again to determine if adequate ventilation is possible. If the amount of ventilation is too high or too low, this will provide feedback to the model of artificial neural networks that the software is hosting, allowing the model parameters to be updated. That is, after the software has decided to keep a window open for a certain amount of time in order to achieve the instantaneous desired amount of humidity and temperature and when the window is closed again, if the measurement determined that the desired humidity and temperature are still not reached, the window will determine the opening time accordingly if it tries to reach the next target humidity and temperature. If more time is needed to reach humidity and temperature, the window will be held open for longer, and the window will be open for shorter periods if less time is needed. The artificial neural network model hosted by the software determines the ideal holding times by continuously performing this measurement.

In other words, it will be calculated how soon after which greenhouse group in the greenhouse group needs aeration for the change in the greenhouse temperature and humidity information presented as input to the algorithm. This calculation will include when the critical threshold for temperature and humidity values will be reached, specific to the product that is attached to the greenhouse. Using wind direction and violence information at the set time, it will also be calculated which window in the greenhouse group must remain open for how long to allow the corresponding greenhouse to be aerated to the prescribed amount. This will close the windows again to measure the temperature and humidity within the greenhouse. In the event that the target value cannot be reached in these measurements, the difference between the target value and the measured value will also be introduced into the system and the calculations will be made again to minimize this difference in the next ventilation. In this way, as the number of vents increases over time, the difference between the target value and the measured value will be closer to zero, and the system will learn the correct ventilation time on its own. All of these calculations and the minimizing process will be performed through an artificial neural pattern.

Example Scenario 1 :

Wind at 3 magnitudes from above reaches the group of greenhouses, which appears in Figure 1. The system is calculated from each greenhouse as humidity, temperature, wind measurement, and opening windows perpendicular to the wind direction, as shown in Figure 2, for best ventilation. Example Scenario 2:

Wind at 2 upwards and 1 in the lower-left direction reaches the greenhouse group, which appears in Figure 3. In this case, the ventilation of greenhouses in the greenhouse group switched to the ventilation pattern shown in Figure 4, with the processing of data from the sensors in the greenhouses.

Although only wind direction and intensity are shown in these scenarios, the data obtained from meteorology are calculated as a preliminary calculation for optimal utilization of the temperature in the greenhouse by using the product type in the greenhouses with wind expectations. The example model shapes and window openings listed above are assumptions. Real greenhouse geometries contain more irregular shapes. With the location of these irregular shaped greenhouses relative to each other, calculating the wind direction, temperature humidity data, the window opening and how open it will remain between the hours is not a calculation that a person can perform in a moment.

Claims

1. An intelligent control system for greenhouse air conditioning ensuring ideal air- conditioning in coordination with the greenhouses in the greenhouse group where several greenhouses are included is characterized in that comprising an algorithm calculating which windows in the greenhouses should be ventilated for how long according to the rate of change in the airflow rate and wind direction information collected from the greenhouses, by comparing the humidity and temperature obtained as a result of ventilation with the necessary humidity and temperature collected from greenhouses as regards temperature, humidity sensors, anemometer and wind station and Internet connection module, by comparing the humidity and temperature obtained as a result of ventilation in the greenhouses, by comparing the humidity obtained as a result of the ventilation and software containing an artificial neural network model that decides whether the ventilated time is sufficient or not during the next ventilation phase accordingly.