Abstract: The idea of irrigation is not new, irrigation stems as far back as the Egyptians and probably further in unrecorded history. Even the idea of automated irrigation is not new, mankind has figured out how to irrigate large areas of foliage through the use of automated and drop irrigation systems. Efficient, automated irrigation systems, which can irrigate plants to a desired level and supply those plants with just the amount of water required for normal an uptake plant growth, are currently not available. These systems, if developed, could reduce waste of irrigated water. The irrigation controller is the "brain" of an entire irrigation system. It supervises the flow of water and fertilizer to the plants, therefore, enables the farmer, or the gardener, to obtain optimized results: A successful crop or a beautiful garden, by using an optimum amount of water and fertilizer. Nowadays computerized control is very essential for the greenhouse irrigation control. Many conventional methods for controlling greenhouse irrigation are not effective since they are either based on on-off control methods or proportional control methods. This results in a loss of energy and productivity. The paper presents a solution for an irrigation controller based on the fuzzy-logicmethodology.59218
First, it describes the general problem of irrigation. Then, it discusses the physical control model. The developed Fuzzy Logic Controller (FLC) prototype is based on a Mamedani controller and it is built on MATLAB software. Following the discussion and the formal presentation of the fuzzy controller, the paper provide examples that will show the simplicity in designing and constructing such a system and other advantages of using fuzzy logic in the feedback control problem. The developed fuzzy logic controller can effectively estimate amount of water uptake of plants in distinct depth using the reliable irrigation model, evapotranspiration functions, environmental conditions of greenhouse, soil type, type of plant and another factors affecting the irrigation of greenhouse.Key words:  Control • Fuzzy • Automation • Greenhouse • Irrigation systemINTRODUCTIONWater is a basic component of all known life on Earth. Water can both sustain life in correct quantities and threaten life when it is not available oroverabundant. Water as a result is a very precious natural resource that must not be wasted. If too much water  is  applied the problems arise consisting ofrunoff,  erosion,  waste  of  water and deceased plant life. If  too  little water is applied different problems arise such as turf burnout. The key in irrigation is striking to correct balance for optimal plant life with optimal use of water [1, 2].Irrigation controllers are pided roughly into two main classes [3]. Open loop controllers: These are based on a pre-defined control concept, with no feedback from thecontrolled object. Most (if not all) of the simplecontrollers operate in this fashion. The user sets the time to start, the time to end, the pause intervals and the watering periods. These parameters are preset for the entire session. That is: • How long the irrigation session should last,•源]自[751^`论\文"网·www.751com.cn/
How often the irrigation period should repeat itself and• How much water (and/or fertilizer) will be used in these irrigation sessions.No checking is done to know whether the right amount of water is used or not. These types ofcontrollers, though relatively cheap, are not very good, since in most cases they do not provide the optimal (or even a good) solution to the irrigation problem. The major factor in the irrigation process is the time.Therefore, the open-loop controller uses a periodicirrigation  policy  [4].  In  this  policy,  the  irrigation  is based on the relevant amounts of water that must be given periodically (a large amount once in few days, or fractions on each day). The experts claim that periodic irrigation with large amounts is better because it washes the soil free of chemicals and creates a better balanced soil chemically [5].Closed-loop controllers: These are based on acombination of pre-defined control concept (feed-forward) and feedback from the controlled object. In this type of controller, there is a feedback of thenecessary data to determine the amount of water needed for irrigation. There are several parameters that should influence the decision of how much water to use in the irrigation process. Some of these parameters are fixed for the session and are of an agricultural nature (such as the kind of plants, kind of soil, leaf coverage, stage of growth, etc.) and some of them vary and should be measured during the irrigation process. Theseparameters are of a physical nature (such astemperature, air humidity, radiation in the ground, soil humidity, etc.). So when these conditions change, the amount of water being used for the irrigation should change also [6].The  system  described  in  this paper utilizes closed-loop control. The controller receives feedback from  one  or  more  sensors in the field, thatcontinuously provide updated data to the controller about parameters that are influenced by the systembehavior  (such  as  soil  moisture  level,  temperaturein hothouses and so on). According to the measurements provided by the sensors and the pre-programmed parameters (such as the kind of plants and the saltiness of the ground), the controller decides on how far to open the water valve. The major parameters that determine the irrigation process are: • Type of growth;• Status of the growth (height, depth of roots);• Leaf coverage;• Kind of soil and saltiness;• Water budget (economy or normal irrigation).Therefore, the input parameters that are used by the system are: • Soil (ground) humidity;• Temperature;• Radiation;• Wind speed;• Air humidity;• Salinity (amount of salt in the ground).The output parameters are: • Opening/closing the valves for water and/orfertilizer and adjusting their amounts incombination;• Turning energy systems on/off (lights, heating,ventilation);• Opening/closing walls and roofs of hothouses [7].DESIGN OF A FUZZY IRRIGATION CONTROLLERFigure 1 depicts the block diagram of the controller embedded in the system model. As can be seen, the controller is operated in four interrelated stages. • Desired soil moisture: This block shows the set point of soil moisture that plant can grow up properly.• The input variables of soil model: In this stage some variables represent influence on the rate of soil evaporation such as: Temperature, airhumidity, wind speed, radiation.• The soil evaporation model stage. This converts the water flow rate, temperature, air humidity, wind speed and radiation to the actual soil moisture• The control stage: In this stage the desires soil moisture is compared with the measured soilmoisture following the comparison, a dynamicdecision is made regarding the amount of water to be added to the soil.In continuation any of four stages will consider that how modeling.Desired soil moisture: At first according to the kind of plant and type of growth extract amount of water that is necessary  for  growth  and  then  with  consideration of Fig. 1: Irrigation controller block diagram and system Model kind of soil calculate desired soil moisture that it's different for any kind of plant, type of growth and kind of soil. An assumed graph of desired soil moisture is shown in Fig. 2.The input variables of soil model: In addition to the amount of water to be added to the soil, four effective factors as: temperature, air humidity, wind speed and radiation influence on the soil evaporation. The input variables were defined as follows:Temperature: This variable should be defined as acontinuous signal (normally as a sine wave whichsimulated the day and night temperature changes), but my show sharp changes in special places like deserts and so on therefore:• A sine wave with amplitude of 5 ºC;• A frequency of 0.2618 rad/h. This frequency is measured according to a time period of 24 h:0.2168 rad/h = 2p/T=2p/24.• A constant bias(offset) of 30 ºC;This stimulus generates a wave which at itsmaximum can reach 35°C (midday) and at its minimum can reach +25°C (midnight). In this way, thetemperature on any given day can be simulated by changing the bias that is attached to the variable. This persion is obtained by uniform number generation (Light red graph in Fig. 3).The Air humidity variable: • A sine wave with amplitude of 10%;• Bias of 60% (constant);• A frequency of 0.2618 rad/h (blow graph in Fig. 3).The wind speed variable: • 大棚自动滴灌系统英文文献和中文翻译:http://www.751com.cn/fanyi/lunwen_64424.html