Особливості побудови нейро-нечіткої системи керування енерговитратами у теплицях

Abstract

DESIGN peculiaritIES of neuro-fuzzy CONTROL SYSTEM OF energy consumption in greenhouses

V. Lysenko, A. Dudnyk, I. Yakymenko

Greenhouses are characterized by the presence of significant energy streams used to provide the appropriate technology. High energy prices (natural gas, electricity) create the conditions for the development of special systems that can reduce, but better minimize energy costs. This is explained by the fact that existing systems, as constituent technologies, implement the simplest stabilization algorithms that do not provide high energy and resource efficiency (information about the states of the biological component of the control object, the results of prediction of natural perturbations affecting the performance of a living organism is not used).

Therefore, the creation of resource-efficient algorithms for controlling energy flows in such objects using the device of neural networks and fuzzy logic is an actual scientific problem.

The results of previous studies [1-5] have allowed us to conclude that additional information on the predicted values of ambient temperature, solar radiation, information on the states of biological content can create a knowledge base and use it to form control impacts on biotechnical objects in order to minimize energy consumption. , while ensuring the production of products of the required quality and volumes.

The implementation of the system involves the use of a large number of equipment, a control function for which the computer, to which controllers are connected, is running. Controllers control the flow of information in two directions: on the computer and from the computer. Information is transmitted to the computer by processor sensors, and information from the computer enters the actuators (fans, heaters, pumps, etc.). The proposed system using neuro-fuzzy methods and software based on the predicted values of external natural disturbances and the current parameters of the technological process will provide support for decision-making, control and monitoring of the parameters of the biotechnical object.

The purpose of the study is to increase energy efficiency of the work of the electrotechnical equipment in greenhouses through the use of a neuro-fuzzy control system.

A system of fuzzy output is a process of obtaining fuzzy conclusions about the required algorithm for managing the object based on fuzzy conditions and prerequisites that represent information about the current state of the object.

The basis of the rules of fuzzy output systems is intended for the formal representation of the empirical knowledge of experts in one or another subject area in the form of fuzzy production rules.

In order to solve the problem of reducing energy costs, it is proposed to construct a fuzzy expert system for determining the inclusion of pump group groups depending on the values of the outside temperature, time of day and given according to the technological requirements of the temperature in the greenhouse space. Fuzzy output system is the process of obtaining fuzzy conclusions on object management based on fuzzy conditions and prerequisites (for example, the uncertainty of the influence of external natural disturbances), which also takes into account the information about the current state of the object. We set the structure of the system of prediction. At the input are three values the value of external temperature (tz), time of day and temperature in the middle of the greenhouse (tv). At the outlet we set heating circuits - I - under tent heating, II - heating in the zone of growth of plants, III - overhead heating.

Research results. The result of the system is displayed in the rules view (Fig. 5). By changing the position of the slider of the input parameters, we obtain the prediction of activating the respective heating circuits.

The results of the expert system are shown in Fig. 6, under initial conditions: the temperature is from below -7 degrees C. During the whole day, the temperature is necessary in the room of the greenhouse of 20 degrees C.

Conclusions. As a result of the research of the neuro-fuzzy control system of operating modes of electrical equipment in greenhouses, it has been established that the energy efficiency of the equipment can be optimized, that is, the introduction of a control system for switching on pumps allows to reduce the cost of electricity by 25-30%.