【正文】 re door and to the inevitable heat exchanges with outdoor air, even when the cold store door is closed, have been realized. These tests have been performed at various temperature levels for the air in the coldstore and, precisely, at 5, 0 and 25 8C, opening the cold store door every 20 min for about 5 min with an outdoor air temperature of about 18 8C. Moreover, in some tests the cooling load has been obtained by means of controllable electrical heaters located in the cold store, while in other tests a real cooling load has been considered represented by 200 kg of fruit and vegetables for whose preservation the temperature has been fixed at 5 8C in the cold store. In these last two Fig. 1. Sketch of the experimental plant. 642 Table 1 Transducers specifications Transducer C. Aprea et al. / International Journal of Refrigeration 27 (2004) 639648 Coriolis effect mass ?ow rate meter RTD 100 4 wires Piezoelectric absolute pressure gauge Wattmeter Electric energy meter Range 02 kg/min 2100 to 500 8C 110 bar。 130 bar 03 kW 360420 V。 016 A Accuracy ^% ^ 8C ^。 ^ ^% ^% .。 ^% . situations, the cold store door has been opened every 10 min to simulate a real working condition。 moreover, the tests have been performed both in the winter and in the summer season. As for the summer tests the outdoor air temperature at the condenser has been kept at about 32 8C thanks to a channel where the air is heated by means of an electric heater, while in winter the outdoor air temperature has been kept at 10 8C. The experimental results are mostly presented in terms of electrical energy consumption, measured by means of an opportune electric energy meter, evaluating the energy saving obtained when a pressor speed control is used. The tests, which lasted 2 days, have been realized for the R407C and the R507. 4. Fuzzy logic in the pressor speed control The fuzzy logic represents a methodology that allows us to obtain defined solutions from vague, ambiguous or uncertain information. For this the fuzzy process is very similar to that of the human mind capable of finding defined conclusions starting from approximated information and data. In contrast to the classic logic approach, that requires an exact definition of the mathematical model equations characterizing the phenomenon, the fuzzy logic allows us to solve problems not well defined and for which it is difficult, or even impossible, to determine an exact mathematical model. Therefore, the human experience and knowledge is necessary for this type of modelling. In particular, the fuzzy logic is a valid alternative for the solution of nonlinear control problems. In fact the nonlinearity is treated by means of rules, membership functions and inferential process, that ensure simpler implementations and minor design costs. On the other side the linear approximation of a nonlinear model is simple enough, but it has the disadvantage to limit the control performances and can result, in some situations, expensive. Moreover, the fuzzy controllers are robust and allow us to realize improvements or changes in a very simple way by means of the use of the other rules or the membership functions. Many examples of fuzzy control can be found in some recent applications. In particular, in the heating ventilation and airconditioning industry there are various fuzzy control applications of the air temperature and humidity [2528]. The design of a fuzzy controller requires three essential phases. The first is to establish the input and output variables. The second is to define the membership functions for the input and output variables. The last is to select or formulate the control rules. The main goal of this paper is to determine a fuzzy controller capable of regulating the pressor electric motor supply current frequency. In Fig. 2 a block diagram of the fuzzy control process of the mercially avalaible cold store air temperature is reported. In particular, the figure shows a twoinput oneoutput fuzzy controller. The input variables are the temperature difference between the set point temperature and the real temperature of the air in the cold store 240。DT222。 and the derivative of this temperature difference with time 240。d240。DT222。=dt222。 the fuzzy output variable is the frequency of the supply current of the pressor electric motor 240。f 222。: The fuzzy logic is based on the determination of the fuzzyset that represents the possible values of the variables. The fuzzy theory with respect to the traditional logic theory, according to which an element can belong or not to a particular set, allows the partial membership of an element to a set. Each value of the variables is characterized by a membership value which changes with continuity from zero to one. Thus, it is possible to define a membership function for each variable that establishes the membership rate of a variable at a certain set. From an operative point of view, a controller fuzzy receives the values of the input variables, performs some operations and determines an output value. This process is character ized by three principal phases: fuzzification, inference mechanism and defuzzification. The fuzzification process allows to transform a value defined into a fuzzy value。 the inference process determines the output fuzzy by means of the rules fixed according to the experimental reality。 the defuzzification process