(340c) Control Strategies for a Vapour Compression Refrigeration System Used in Mango Exports: An Alternative to Traditional on-Off Controllers

Tropical fruits are important products on the global market. The change to a healthier nutrition, the development of new products and great availability led to a rise in their consumption during the last decade [1-2]. Due to their perishable nature they are stored at lower temperature. To achieve a rapid and efficient decrease in product temperature, refrigeration systems are employed, using vapor compression refrigeration plants. They consist of four main components: the compressor, the condenser, the expansion valve and the evaporator. Within the system a refrigerant is circulating. Though designed to satisfy maximum load, these plants usually work at part-load for much of their life, generally regulated by on/off cycles of the compressor, working at nominal frequency of 50 Hz [3]. The high cost involved in developing cold storage or controlled atmosphere storage is a pressing problem in several developing countries [4].

This contribution presents development and comparison of various strategies for the control of a refrigeration plant used for fruit cooling. The starting point is a model of the plant which is able to simulate both the chamber and the fruit temperature. The model is based on energy balances for each section of the refrigeration system and the fruits. The model is the basis for various control strategies capable of regulating continuously the compressor’s frequency to achieve the desired temperature:

• A PI controller which uses the difference between the desired temperature and the actual temperature as input. Additionally, the proposed PI controller has anti reset windup mechanisms and noise filter at the output. The parameters were tuned using the Ziegler Nichols method and the mathematical model of the refrigeration plant;
• A fuzzy controller, which considers the difference between the desired temperature and the actual temperature, as well as the derivative of the error as inputs. To model relationships between the compressor’s frequency and the chamber temperature, as well as between the temperature of the different sections of the refrigeration chamber, membership functions are defined based on previous knowledge and experimental investigation.

The control strategies are then implemented in an experimental refrigeration plant located on the campus of the University of Piura (Peru). The temperature predictions are analysed for various amounts of mangoes stored in the plant, while the fuzzy and the PI controllers are compared with the conventional on/off approach. The model is able to accurately predict the chamber temperature profiles, but is more sensitive when simulating the fruit temperature. The fuzzy controller is able to achieve the set-point more accurately and in a shorter time than both the PI or the on/off control, achieving a decrease in energy consumption as well.

1. Provido, 2016, Business Mirror, https://businessmirror.com.ph/2016/09/06/global-demand-for-tropical-fru…
2. Altendorf, 2017, Developments in bananas and major tropical fruits, pp. 1-2
3. Aprea, et al., 2004, International Journal of Refrigeration 27, pp. 639-648
4. Basediya, et al., 2013, J Food Sci Technol 50 (3), pp. 429-442