Unicamp - Press Room

Engineer creates automated refrigeration system

Chemical engineer Flávio Vasconcelos da Silva: strict temperature control

After four years of research, chemical engineer Flávio Vasconcelos da Silva, using fuzzy logic, managed to build a prototype of a fully automated refrigeration system, based on the development of artificial intelligent techniques. This is the first work in Brazil, developed in the laboratories of a university, focused on food preservation.

The researcher explains that the automated refrigeration process is intended both for cooling juices or any type of liquids, as well as for the pasteurization of dairy products, for which there is a combination of heating and cooling, which occurs extremely quickly, causing a thermal shock of approximately 70 degrees. As for rapid cooling, there must be strict temperature control of the cold fluid (water or propylene glycol, a non-toxic antifreeze), to maintain ideal pasteurization conditions.

In many processes, temperature controls are required that are compatible with the conditions of the product that needs to be stored. For example: “When you intend to pasteurize products derived from milk, strict control of the temperatures of the cold fluid is necessary, essential to annihilate certain organisms that are harmful to health, without altering the flavor and quality of products such as milk, cheese, yogurt, beer and wine”, explains the engineer.
According to Flávio, the idea is to develop this technology in Brazil, although, obviously, this requires more in-depth studies. “It is after the assembly process and complete instrumentation of the prototype that our equipment, called chiller, is able to carry out perfect 'supervision' of the refrigeration system and control temperatures in the processing of juices, milk and cheese, among others. other products”, says the engineer.

Artificial intelligence – Flávio Vasconcelos is the author of the doctoral thesis Comparison of the performance of a refrigeration system for liquid cooling controlled by different control modes, to be defended on the 13th, under the guidance of professor Vivaldo Silveira Júnior, from the Faculty of Food Engineering ( FEA).

The researcher explains that the design of this equipment is basically intended for the food preservation industry, but that it can also be a system aimed at other types of industries, such as chemistry and petrochemistry. In a refrigeration cycle, the efficiency of the system is directly linked to the ability to maintain temperatures and pressures corresponding to those required for processing different types of food. Depending on the product under development, the temperatures of the refrigeration system have a great influence on energy consumption and general performance of the process as it is highly influenced by climatic variations and processing changes. Controlling the temperature of the liquid cooled in the chiller is directly linked to the sensitivity of the product placed to be cooled.

In Brazil, the use of fuzzy control in industrial processes, although it has been growing rapidly, is still a very recent process. The development of artificial intelligence techniques in recent years, according to Flávio, is increasingly occupying prominent positions in research in the area of industrial process control and, little by little, is beginning to be implemented in industrial complexes with enormous success.

Most of the work found in the literature, even at an international level, is based on process simulations. “Practical work like the one we are developing here in the FEA laboratories at Unicamp is still rare. And in complex systems, such as those relating to refrigeration systems, they are even scarcer”, observes the FEA researcher.
There are currently some research groups in the fuzzy area in Brazil, but curiously none in the food area. There are no companies in the country investing in fuzzy, “except those that develop small technologies, such as a refrigerator with fuzzy logic, but it is still an imported technology”.

As for companies, some of them are already starting to invest in fuzzy control in the refrigeration area. In particular, compressor manufacturers, who use a technology currently only developed in countries in Europe, the United States and Japan, according to Flávio.

------------------------------------------------

WHAT IS FUZZY LOGIC

Fuzzy logic is a technique that incorporates the human way of thinking into a control system. A typical fuzzy controller can be designed to behave according to deductive reasoning, that is, the process that people use to induce conclusions based on information they already know. Human operators, for example, can control industrial processes and plants with non-linear characteristics and even little-known dynamic behavior, through experiments and inferences of relationships between process variables. Fuzzy logic can capture this knowledge in a controller, enabling the implementation of a computational controller with performance equivalent to a human operator.

------------------------------------------------