Adaptive control system for controlling thermal comfort within an environment and method therefor

Description

[0001] The present invention relates to an adaptive control system for controlling the thermal comfort inside an environment.

[0002] In particular, said environment can be a working environment or a domestic environment and the present invention refers to an adaptive control system designed to control the level of thermal comfort based on the needs of a user positioned within said environment.

[0003] More specifically, the control of the level of thermal comfort perceived by a user inside said environment is based on the acquisition and processing of a plurality of data (for example data relating to the internal temperature of said environment and to the temperature variation) and based on parameters set by the user.Prior art

[0004] The expression "thermal comfort" means "the psychophysical state in which a subject expresses satisfaction with the thermal environment", i.e. said subject has neither the sensation of heat nor the sensation of cold.

[0005] This psychophysical state is fundamental for a person's health and personal satisfaction can lead to well-being in the home, as well as better productivity in a work environment.

[0006] For practical reasons, in winter it is advisable to set the internal temperature between 21 °C and 23 °C, while in summer it is advisable to set a higher temperature than the one set for the winter to reduce the passage of heat from the outside towards the internal in order to save energy.

[0007] These temperature settings have been confirmed to meet the needs of 80% of people and are recommended by the American Society of Heating Refrigerating and Air Conditioning Engineers (ASHRAE).

[0008] The control of HVAC (Heating, Ventilation and Air Conditioning) systems is directed to improve the quality of an indoor environment, increasing thermal comfort and reducing energy consumption.

[0009] The use of ON-OFF controllers or the use of PID controllers is known to regulate the operations of an HVAC system.

[0010] The ON-OFF regulators are intuitive and simple and regulate a temperature differential between two predetermined values with respect to a desired internal temperature of the environment: a predetermined first value and a predetermined second value, greater than said predetermined first value.

[0011] However, a disadvantage of the use of these ON-OFF regulators is due to the fact that the thermal comfort control through said ON-OFF regulators is subject to a wide fluctuation in temperature.

[0012] Moreover, a further disadvantage is given by the fact that it is not possible to control the thermal comfort with time delays by means of these ON-OFF regulators.

[0013] These ON-OFF controllers regulate thermal comfort using the internal temperature measured moment by moment by temperature sensors placed in the environment, without taking into account additional information, such as climate change or the heat exchange of the building.

[0014] The use of PID controllers (Proportional-Integrative-Derivative) allows an accurate thermal comfort control, using a dynamic error related to the temperature, which is the variable to be controlled.

[0015] However, a disadvantage is that it is necessary that said PID controllers are frequently calibrated with respect to the environmental conditions.

[0016] A common disadvantage to the use of ON-OFF controllers and PID controllers is that, due to the ability of a building to store heat, the HVAC systems controlled by the aforementioned controllers respond to changes in internal temperature with considerable delays.

[0017] This can cause a cooling or overheating of the environment of a building, and consequently a higher energy consumption.

[0018] Due to the different influences on the general quality of the indoor environment, the control system of an HVAC must correlate input variables to output variables.

[0019] For this reason, the control system has numerous elements of uncertainty.

[0020] A further known type control system for controlling thermal comfort comprises one or more microclimatic control units, each of which is able to acquire a plurality of environmental parameters (such as, for example, the air temperature, the surface temperature, the temperature of liquids and fumes, the radiant temperature, the relative humidity, the atmospheric and differential pressure, the air velocity, the illumination, the solar radiation and the radiant asymmetry, the thermal flow, the gas concentration and any meteorological quantities) and a logic control unit configured to apply subjective parameters, and to provide as output the value of a PMV (Predicted Mean Vote) index concerning the position in which said microclimatic control units are installed.

[0021] The PMV index is one of the main indicators according to the UNI EN ISO 7730 standard of the thermal comfort of an indoor environment.

[0022] This index predicts the average response (in statistical terms) of the thermal sensation of a large part of people exposed to certain thermal conditions for long time periods.

[0023] The classic formula with which it is possible to calculate the PMV index was presented by Fanger in 1972 and depends on six variables: i. the metabolic rate, ii. the insulation provided by clothes, iii. the temperature of the indoor environment, iv. the humidity of the indoor environment, v. the speed of the air, vi. the mean radiant temperature.

[0024] According to a sensitivity analysis published in G. M. Revel, E. Sabbatini, and M. Arnesano, "Development and experimental evaluation of a thermography measurement system for real-time monitoring of comfort and heat rate exchange in the built environment" Measurement Science and Technology , vol. 23, no. 3, 2012., the most influential variables are the metabolic rate, the temperature of the indoor environment and the humidity of the indoor environment.

[0025] Without loss of generality, human activities and variables related to clothing can be estimated based on the ISO-7730 standard.

[0026] According to these hypotheses, the final formula of PMV can be a nonlinear function of measurable environmental variables (at least temperature and humidity).

[0027] Currently, the known control systems through which it is possible to manage the level of thermal comfort mainly use control techniques based on the use of PID controllers or control techniques aimed at optimally limiting energy consumption.

[0028] Some known control systems may comprise one or more NEST thermostats, i.e. a programmable thermostat that requires the configuration of parameters based on user preferences, or one or more KNX thermostats that require data from devices / systems that use the same communication protocol.

[0029] Patent document US-2019 / 353377-A1 describes a control system adapted for controlling at least one air conditioning unit and comprising temperature sensors, a user interface module and a control unit to acquire temperature data.Aim of the invention

[0030] The aim of the present invention is to overcome said disadvantages by providing an adaptive control system to control thermal comfort in an environment, in which said control system has a simple structure and operation (with respect to the complexity of the functions used and the calculations that the control systems of known types must be carried out to regulate thermal comfort) and is able to control thermal comfort in a fine manner.

[0031] A second aim of the present invention is to provide a control system for controlling thermal comfort, also reducing energy consumption.

[0032] A further aim of the present invention is to provide a versatile control system both for a user (so that the user can preferably be an active part in choosing the level of thermal comfort) and for an operator. In fact, this control system offers a user the possibility of customizing some parameters and an operator the possibility of carrying out maintenance in a simple way.Object of the invention

[0033] It is therefore object of the invention an adaptive control system according to claim 1.

[0034] Further embodiments of the control system are disclosed in the dependent claims.

[0035] It is object of the invention also an air conditioning unit comprising said adaptive control system.

[0036] It is object of the invention also a method for controlling the thermal comfort inside an environment through said control system.Figure list

[0037] The present invention will be now described, for illustrative, but not limitative purposes, according to its embodiment, making particular reference to the enclosed figures, wherein: Figure 1 is a schematic view of an adaptive control system for controlling the thermal comfort level of an environment according to the invention, wherein said control system comprises a control logic unit and a user interface module associated with an air conditioning unit arranged in said environment, in which the logic control unit is configured to communicate with said user interface module and with said air conditioning unit; Figure 2 is a schematic view of a first part of a neural network showing four input nodes and some groups of output nodes; Figure 3 is a schematic view of a second part of the neural network showing the remaining groups of output nodes (and the input nodes shown in Figure 2 are again shown). Detailed description of the invention

[0038] With reference to Figure 1, an adaptive control system for controlling the thermal comfort inside an environment.

[0039] In the embodiment being disclosed, an air conditioning unit U is arranged inside said environment.

[0040] An air conditioning unit U comprising a fan F, adjusting means A for adjusting the speed of said fan F, control valve V for controlling the flow rate of a thermovector fluid over time.

[0041] The control system comprises: temperature detecting means 1, to be installed in said environment, configured to detect a respective value of internal temperature T INi of said environment for a plurality of predetermined time periods Δt REFi , with i=1,...,N, wherein N is a positive integer, a user interface module 2 configured to allow a user to select at least one temperature value T U and a speed value V U of the fan F, storage means 3, and a logic control unit 4 connected to said temperature detecting means 1, to said user interface module 2, and to said storage means 3.

[0042] In the embodiment being disclosed, said temperature detecting means 1 comprise a temperature sensor.

[0043] Furthermore, said storage means 3 comprise a memory.

[0044] In particular, the memory is arranged inside said logic control unit 4.

[0045] However, it is not necessary that the memory is arranged inside said logic control unit 4.

[0046] Furthermore, the logic control unit is configured to: ∘ acquire a value of internal temperature T INi from said temperature detecting means 1 for a respective predetermined time period Δt REFi , as well as a temperature value T U and a speed value V U from said user interface module 2, ∘ calculate one or more values of temperature variations δT Dj , with j=1,...,M, wherein M is a positive integer and j is the number of predetermined time periods, and wherein each value of temperature variation δT Dj is obtained by the subtraction between the value of internal temperature T INi in a predetermined time period and the value of internal temperature T INi-1 in the previous predetermined time period; ∘ store in said storage means 3: the values of internal temperature T INi referred to at least two predetermined time periods Δt REFi and said one or more values of temperature variation δT Dj , as well as said temperature value T U and said speed value V U , a first group of fuzzy sets F1 comprising a plurality of fuzzy sets F1 1 ,F1 2 ...F1 N (i.e. at least two fuzzy sets), each of which identifies a respective degree of belonging of a value of internal temperature T INi , a second group of fuzzy sets F2 comprising a plurality of fuzzy sets F2 1 ,F2 2 ...F2 N (i.e. at least two fuzzy sets), each of which identifies a respective degree of belonging of a value of temperature variation δT Dj , a third group of fuzzy sets F3 comprising a plurality of fuzzy sets F3 1 ,F3 2 ...F3 N (i.e. at least two fuzzy sets), each of which identifies a respective degree of belonging of an opening state of the control valve V, a fourth group of fuzzy sets F4 comprising a plurality of fuzzy sets F4 1 ,F4 2 ...F4 N (i.e. at least two fuzzy sets) each of which identifies a respective degree of belonging of a value of the speed of the fan F, a plurality of predetermined inferential rules that establish on the basis of the degree of belonging of each value of internal temperature T INi and the degree of belonging of each value of temperature variation δT Dj , the degree of belonging of the opening state of the control valve V and the degree of belonging of the speed of the fan F.

[0047] Each group of fuzzy sets F1, F2, F3, F4 comprises a respective a plurality of fuzzy sets, particularly at least two fuzzy sets.

[0048] Each fuzzy set identifies a degree of belonging (in other words each fuzzy set is associated with a degree of belonging).

[0049] Consequently more fuzzy sets (and then more degrees of belonging) can be associated with a same value of an input variable which can be a value of internal temperature T INi , a value of temperature variation δT Dj , a value associated with to an open state of the control valve V and a value associated with the speed of the fan F.

[0050] This is due to the fact that, when a fuzzy domain is used, a value of an input variable can belong with a degree of belonging to one or more predetermined classes which are established based on type of the application.

[0051] For example, a temperature value equal to 25°C can belong to a class "warm" which has a degree of belonging equal to 0,8 and to a class "cold" which has a degree of belonging equal to 0,4.

[0052] In this example, two fuzzy data (0,8 and 0,4) are associated with to the same temperature value (25°C) which is a precise data.

[0053] The inferential rules are established a priori, i.e. before the use of the control system.

[0054] Said inferential rules allow to establish the degree of belonging of the opening state of the control valve V to each fuzzy set F3 1 ,F3 2 ...F3 N of the third group of fuzzy sets F3, and the degree of the belonging of the speed of fan F to each fuzzy set F4 1 ,F4 2 ...F4 N of the fourth group of fuzzy sets F4, on the basis of the degree of belonging to each value of internal temperature T INi and of the degree of belonging to each value of variation temperature δT Dj .

[0055] Furthermore, the logic control unit 4 is configured to: ∘ perform a respective fuzzification (or fuzzy transform) referred to: each value of internal temperature T INi to obtain a respective first fuzzy data set P1 i comprising a respective plurality of first fuzzy data P1 i1 ,P1 i2 ... P1 iN , wherein each first fuzzy data identifies a respective degree of belonging of said value of internal temperature T INí , and is obtained by means of a predetermined first equation defined by a plurality of parameters, each value of temperature variation δT Dj to obtain a respective second fuzzy data set P2 i comprising a respective plurality of second fuzzy data P2 i1 ,P2 i2 ...P2 iN , wherein each second fuzzy data identifies a respective degree of belonging of said value of temperature variation δT Dj , and is obtained by means of a predetermined second equation defined by a plurality of further parameters, ∘ apply said predetermined inferential rules to each pair of fuzzy data formed by a first fuzzy data P1 i1 ,P1 i2... P1 iN and a second fuzzy data P2 i1 ,P2 i2 ...P2 iN to obtain a respective third fuzzy data set P3 i comprising a plurality of third fuzzy data P3 i1 ,P3 i2 ...P3 iN , wherein each third fuzzy data identifies a respective degree of belonging of said opening state of the control valve V with a respective fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3 i , and a respective fourth fuzzy data set P4 i comprising a plurality of fourth fuzzy data P4 i1 ,P4 i2 ...P4 iN , wherein each fourth fuzzy data identifies a respective degree of belonging of said value of the speed of the fan F with respect to a respective further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4 i , ∘ perform a respective defuzzification (or fuzzy inverse transform) referred to: said fuzzy data set associated with to said third fuzzy data of said third fuzzy data set P3 i to obtain a first real number R1 indicating if the control valve V is to be open and how much to open or if it is to be closed, said further fuzzy data set associated with fourth fuzzy data of said fourth fuzzy data set P4 i to obtain a second real number R2 indicating the speed of the fan F.

[0056] The fuzzification allows precise numeric data, i.e. the value of internal temperature T INi and the value of the temperature variation δT Dj , to pass to the fuzzy domain and therefore said precise numeric data become fuzzy data.

[0057] For fuzzification it is possible to use any equation that represents any function adapted to identify a degree of belonging to a fuzzy set.

[0058] For the fuzzification of each value of internal temperature value T INi it is possible to use a predetermined first equation which represents any function identifying a degree of belonging to a fuzzy set of the first group of fuzzy sets.

[0059] For the fuzzification of each value of temperature variation value δT Dj it is possible to use a predetermined second equation which represents any function identifying a degree of belonging to a fuzzy set of the second group of fuzzy sets.

[0060] In particular, in the embodiment being described, with reference to the fuzzification referred to each value of internal temperature T INi , said predetermined first equation is defined by a first triad of parameters a 1 , b 1 , c 1 and represents a Gaussian curve: P 1 i = a 1 ⋅ e − x i − b 1 2 c 1 2 wherein the parameter a 1 is a predetermined number greater than 0 and less than or equal to 1 and represents the height of the Gaussian curve; the parameter b 1 is a predetermined value of temperature in which the Gaussian curve is centred; the parameter c 1 is a predetermined value indicating the distance between a symmetry axis of the Gaussian curve and an inflection point of the Gaussian curve; and x i is a respective value of internal temperature T INi .

[0061] Furthermore, with reference to the fuzzification referred to each value of the temperature variation δT Dj , said predetermined second equation is defined by a second triad of parameters a 2 , b 2 , c 2 and represents a Gaussian curve: P 2 i = a 2 ⋅ e − x j − b 2 2 c 2 2 wherein the parameter a 2 is a predetermined number greater than 0 and less than or equal to 1 and represents the height of the Gaussian curve; the parameter b 2 is a predetermined value of temperature in which the Gaussian curve is centred; the parameter c 2 is a predetermined value indicating the distance between a symmetry axis of the Gaussian curve and an inflection point of the Gaussian curve; and x j is a respective value of temperature variation δT Dj .

[0062] The defuzzification allows the transition from the fuzzy domain to the domain of real numbers.

[0063] The defuzzification is applied on each fuzzy data set of a third fuzzy data of a third fuzzy data set P3 i and on each further fuzzy data set of a third fuzzy data of a fourth fuzzy data set P4 i .

[0064] Said third fuzzy data set P3 i and said fourth fuzzy data set P4 i are obtained through the application of the predetermined inferential rules (stored in said storage means 3) to each possible pair of fuzzy data formed by a first fuzzy data P1 i1 ,P1 i2 ...P1 iN and by a second fuzzy data P2 i1 ,P2 i2 ...P2 iN .

[0065] Said first fuzzy data P1i 1 ,P1 i2 ...P1 iN and said second fuzzy data P2 i1 ,P2 i2 ...P2 iN belong respectively to a first fuzzy data set P1 i and to a second fuzzy data set P2 i , wherein said first fuzzy data set P1 i and said second fuzzy data set P2 i can comprise a respective plurality of fuzzy data.

[0066] The defuzzification can be performed by means of one of the following methods of known type: mean of maxima, last of maxima, center of sums, center of area, center of inertia, center of gravity.

[0067] In the embodiment being disclosed, the defuzzication referred to said fuzzy data set associated with to a third fuzzy data of a third fuzzy data set P3 i is performed through a method called mean of maxima.

[0068] The logic control unit 4 is configured to: ∘ select a maximum value P3 iMAX with respect to each third fuzzy data P3 i1 ,P3 i2 ...P3 iN , wherein said maximum value is associated with at least one fuzzy set F3 1 ,F3 2 ...F3 N of the third group of fuzzy sets F3, ∘ calculate at least two real numbers P3 1 , P3 2 applying a predetermined third equation defined by a third triad of parameters a 3 , b 3 , c 3 : x 2 − 2 b 3 x + b 3 2 + c 3 2 ln P 3 iMAX a 3 = 0 wherein the parameter a 3 is a number greater than 0 and less than or equal to 1 and represents the height of a Gaussian curve; the parameter b 3 is a value of temperature in which a Gaussian curve is centred; the parameter c 3 is the value of the distance between a symmetry axis of a Gaussian curve and an inflection point of said Gaussian curve; and x is the value that P3 can assume; ∘ calculate said first real number R1 equal to the mean value of said at least two real numbers P3 1 , P3 2 .

[0069] Also the defuzzification referred to said further set associated with a fourth fuzzy data of a fourth fuzzy data set P4 i is performed through the method called mean of maxima.

[0070] The logic control unit 4 is configured to: ∘ select a maximum value P4 iMAX with respect to each fourth fuzzy data P4 i1 ,P4 i2 ...P4 iN , wherein said maximum value is associated with at least one fuzzy set F4 1 ,F4 2 ...F4 N of the fourth group of fuzzy sets F4, ∘ calculate at least two reals number P4 1 , P4 2 applying the following predetermined fourth equation defined by a forth triad of parameters a 4 , b 4 , c 4 : x 2 − 2 b 4 x + b 4 2 + c 4 2 ln P 4 iMAX a 4 = 0 wherein the parameter a 4 is a number greater than 0 and less than or equal to 1 and represents the height of a Gaussian curve; the parameter b 4 is a value of temperature in which a Gaussian curve is centred; the parameter c 4 is the value of the distance between a symmetry axis of a Gaussian curve and an inflection point of said Gaussian curve; and x is the value that P4 can assume; ∘ calculate said second real number R2 equal to the mean value of said at least two real numbers P4 1 , P4 2 .

[0071] Furthermore, the logic control unit 4 is configured to perform a neural network (shown in Figures 3 and 4).

[0072] Said neural network comprises an input layer and an output layer.

[0073] In the embodiment being disclosed, the neural network does not have any intermediate layer between the input layer and the output layer.

[0074] The input layer comprises a plurality of input nodes for receiving a respective value.

[0075] In the embodiment being disclosed, the input layer comprises four input nodes: a first input node N IN1 , a second input node N IN2 , a third input node N IN3 and a fourth input node N IN4 .

[0076] In particular, said neural network is a Kohonen network and said first input node N IN1 receives a value of internal temperature T INi , said second input node N IN2 receives a value of temperature T U selected by the user, said third input node N IN3 receives a value of speed V U of the fan F selected by the user and said fourth input node N IN4 receives a value of temperature variation δT Dj .

[0077] The output layer comprises: ▪ a respective first group of output nodes N OUT1F11 , N OUT2F11 ...N OUTNF11 ; N OUT1F12 ,N OUT2F12 ...N OUTNF12 ;... N OUT1F1N ,N OUT2F1N ...N OUTNF1N for each fuzzy set F1 1 ,F1 2 ...F1 N of the first group of fuzzy sets F1, ▪ a respective second group of output nodes N OUT1F21 ,N OUT2F21 ...N OUTNF21 ; N OUT1F22 , N OUT2F22 ...N OUTNF22 ;... N OUT1F2N , N OUT2F2N ...N OUTNF2N for each fuzzy set F2 1 ,F2 2 ...F2 N of the second group of fuzzy sets F2.

[0078] In other words, for each fuzzy set of the first group of fuzzy sets F1 there is a respective first group of output nodes and for each fuzzy set of the second group of fuzzy sets F2 there is a respective second group of output nodes.

[0079] Each first group of output nodes comprises one or more output nodes and each second group of output nodes comprises one or more output nodes.

[0080] In particular, said neural network is configured to: receive over time said value of internal temperature T INi , said value of temperature variation δT Dj , said temperature value T U and said speed value Vu at a respective input layer N IN1 ,N IN2 ,N IN3 ,N IN4 , and provide as output the updated parameters for said predetermined first equation at a respective output node of each first group of output nodes, and the further updated parameters for said predetermined second equation at a respective output node of each second group of output nodes.

[0081] The logic control unit 4 is configured to: ∘ insert said updated parameters in said predetermined first equation to obtain a first updated fuzzy data P1 i and said further updated parameters in said predetermined second equation to obtain a second updated fuzzy data P2 i , ∘ apply said predetermined inferential rules to each pair of fuzzy data formed by a first updated fuzzy data P1 i and a second updated fuzzy data P2 i , so as to obtain a third updated fuzzy data P3 i associated with the control valve V and a fourth updated fuzzy data P4 i associated with the speed of the fan F, as well as to obtain an updated first real number R1 and an updated second real number R2, through defuzzification; ∘ generate a signal to control said control valve V containing an information concerning said first real number R1 or said updated first real number R1 for opening / closing the control valve V and a further signal to control said adjusting means A containing an information concerning said second real number R2 or said updated second real number R2 for the speed of the fan F.

[0082] In the embodiment being disclosed, each first group of output nodes comprises three respective output nodes N OUT1F11 ,N OUT2F11 ,N OUT3F11 , N OUT1F12 ,N OUT2F12 ,N OUT3F12 ,... N OUT1F1N ,N OUT2F1N ,N OUT3F1N and said neural network is configured to provide at each output node of said first group of output nodes a respective updated parameter a 1 , b 1 , c 1 .

[0083] In this way, it is possible to obtain a respective first triad of updated parameters a 1 , b 1 , c 1 .

[0084] Each second group of output nodes comprises three respective output nodes N OUT1F21 ,N OUT2F21 ,N OUT3F21 , N OUT1F21 ,N OUT2F21 ,N OUT3F21 , ... N OUT1F2N ,N OUT2F2N ,N OUT3F2N and said neural network is configured to provide at each output node of said second group of output nodes a respective updated parameter a 2 , b 2 , c 2 . In this way it is possible to obtain a second triad of updated parameters a 2 , b 2 , c 2 .

[0085] In particular, said neural network is a Kohonen network and said input node N IN4 for receiving a value of temperature T U selected by the user and said input node N IN4 for receiving a value of speed V U of the fan F selected by the user are arranged between said input node N IN1 for receiving a value of internal temperature T INi and said input node N IN4 for receiving a value of temperature variation δT Dj .

[0086] Consequently, said logic control unit 4 is configured to insert the first triad of updated parameters a 1 , b 1 , c 1 in the predetermined first equation to obtain a first updated fuzzy data P1 i and to insert the second triad of updated parameters a 2 , b 2 , c 2 in the predetermined second equation to obtain a second updated fuzzy data P2 i .

[0087] In other words, through the neural network it is possible to take into account to control the thermal comfort the temperature T U desired by the user and the speed V U of the fan desired by the user and update both the parameters which are used in the first equation for the fuzzification of the values of internal temperature T INi and the parameters which are used in the second equation for the fuzzification of the values of temperature variation δT Dj .

[0088] Consequently, if at the beginning the fuzzification of the values of internal temperature T INi and the fuzzification of the values of temperature variation δT Dj are respectively obtained through equations which are predetermined (i.e. the predetermined first equation and the predetermined second equation) since the parameters used in the respective equations are decided a priori, through the neural network it is possible to update the parameters present in said equations to adapt the thermal comfort to the needs of the user.

[0089] The output layer of the neural network can comprise: ▪ a respective third group of output nodes N OUT1F31 , N OUT2F31 ...N OUTNF31 , N OUT1F32 ,N OUT2F32 ...N OUTNF32 , ... N OUT1F3N ,N OUT2F3N ...N OUTNF3N for each fuzzy set F31,F32...F3N of the third group of fuzzy set F3, ▪ a respective fourth group of output nodes N OUT1F41 ,N OUT2F41 ...N OUTNF41 , N OUT1F42 , N OUT2F42 ...N OUTNF42 , ...N OUT1F4N ,N OUT2F4N ...N OUTNF4N for each fuzzy set F4 1 ,F4 2 ...F4 N of the fourth group of fuzzy set F4.

[0090] The neural network is configured to provide as output: the updated parameters a 3 , b 3 , c 3 of said third triad of parameters for said predetermined third equation, each of which is at a respective output node of each third group of output nodes, the updated parameters a 4 , b 4 , c 4 of said fourth triad of parameters for said predetermined fourth equation, each of which is at an output node of each fourth group of output nodes.

[0091] The logic control unit 4 can be configured to: ∘ insert said third triad of updated parameters a 3 , b 3 , c 3 in said predetermined third equation to obtain an updated fuzzy data set, due to the change of the shape of the function represented by said predetermined third equation, wherein said updated fuzzy data set is associated with said third fuzzy data of said third fuzzy data set P3 i , ∘ insert said fourth triad of updated parameters a 4 , b 4 , c 4 in said predetermined fourth equation to obtain a further updated fuzzy data set, due to the shape of the function represented by said predetermined fourth equation, wherein said further fuzzy data set is associated with said fourth fuzzy data of said fourth fuzzy data set P3 i , ∘ defuzzify said updated fuzzy data (to obtain a first updated real number R1) and said further updated fuzzy data set (to obtain a second updated real number R2).

[0092] Each third group of output nodes N OUT1F31 , N OUT2F31 ...N OUTNF31 , N OUT1F32 ,N OUT2F32 ...N OUTNF32 ,... N OUT1F3N ,N OUT2F3N ...N OUTNF3N comprises three output nodes N OUT1F31 , N OUT2F31 ,N OUT3F31 , N OUT1F32 ,N OUT2F32 ,N OUT3F32 ....N OUT1F3N ,N OUT2F3N ,N OUT3F3N and said neural network can be configured to provide at each output node of said third group of output nodes a respective updated parameter a 3 , b 3 , c 3 for said predetermined third equation.

[0093] In this way it is possible to obtain a respective third triad of updated parameters a 3 , b 3 , c 3 for each third group of output nodes.

[0094] Said logic control unit 4 can be configured to insert said updated parameters a 3 , b 3 , c 3 in said predetermined third equation to obtain an updated fuzzy data set, due to the change of the shape of the function represented by said predetermined third equation, wherein said updated fuzzy data set is associated with said third fuzzy data of said third fuzzy data set P3 i .

[0095] Each fourth group of output nodes N OUT1F41 , N OUT2F41 ...N OUTNF41 ;N OUT1F42 ,N OUT2F42 ...N OUTNF42 ,... N OUT1F4N ,N OUT2F4N ...N OUTNF4N comprises three respective output nodes N OUT1F41 , N OUT2F41 ,N OUT3F41 , N OUT1F42 ,N OUT2F42 ..N OUTNF42 , ... N OUT1F4N ,N OUT2F4N ... N OUTNF4N , at which a respective updated parameter a 4 , b 4 , c 4 is present.

[0096] In this way it is possible to obtain a respective fourth triad of updated parameters a 4 , b 4 , c 4 for each fourth group of output nodes.

[0097] Said logic control unit 4 can be configured to insert said updated parameters a 4 , b 4 , c 4 in said predetermined fourth equation to obtain a further updated fuzzy data set, due to the change of the shape of the function represented by said predetermined fourth equation, wherein said further updated fuzzy data set is associated with said fourth fuzzy data of said fourth fuzzy data set P4 i .

[0098] In other words, at the beginning, the control logic unit 4 performs the defuzzification of said fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3 i to obtain a first updated real number R1 indicating if the control valve V is to be open and how or if is to be closed, as well as the defuzzification of said further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4 i to obtain a second updated real number R2 indicating the speed of the fan F.

[0099] Subsequently, every time that the parameters of the predetermined third equation are updated, the form of the function represented by said predetermined third equation changes according to said parameters and therefore the fuzzy data set associated with said third fuzzy data of said third data set fuzzy P3 i changes.

[0100] Similarly, every time that the parameters of the predetermined fourth equation are updated, the form of the function represented by said predetermined fourth equation changes according to said parameters and therefore the further set of fuzzy data associated with said fourth fuzzy data of said fourth set of fuzzy data P4 i changes.

[0101] Consequently, every time that the parameters of the predetermined third equation are updated a first updated real number is obtained through the fuzzification of the respective fuzzy data set and every time that the parameters of the predetermined fourth equation are updated a second updated real number is obtained through the defuzzification of the respective further fuzzy data set.

[0102] Finally, the logic control unit 4 generates a signal adapted to control the control valve V containing an information (i.e. said first real number R1) concerning the opening / closing of said control valve V and a further signal adapted to control said adjusting means A containing an information (i.e. said second real number R2) concerning the speed of the fan.

[0103] In other words, at the beginning the logic control unit 4 generates a signal containing a first real number R1 for controlling the opening / closing of the control valve V and subsequently generates a signal containing a first updated real number R1 every time that a value of temperature T U or a value of speed V U of the fan F is selected by an user.

[0104] Therefore, the information contained in the signal generated by the logic control unit 4 is at the beginning a first real number R1 and subsequently a first updated real number R1.

[0105] Similarly to what has been said for the signal, at the beginning the logic control unit 4 generates a further signal containing a second real number R2 for controlling the speed of the fan F through said adjusting means A and subsequently generates a further signal containing a second updated real number R2 every time that a value of temperature T U or a value of speed V U of the fan F is selected by an user.

[0106] Therefore, the information contained in the further signal generated by the logic control unit 4 is at the beginning the second real number R2 and subsequently the second updated real number R2.

[0107] In fact, the update of the first real number R1 and the second real number R2 (and then of the respective fuzzy data sets from which the real numbers mentioned above are obtained) is due to the fact that the logic control unit 4 takes into account of the changes of said value of temperature T U or said value of speed V U .

[0108] After generating the signal and the further signal, the logic control unit 4 will send said signal and said further signal to the air conditioning unit U.

[0109] With reference to the logic control unit 4 disclosed above adapted to perform a fuzzification and a defuzzification as well as a neural network, said logic control unit 4 is a neuro-fuzzy controller.

[0110] As already said, the air conditioning unit U is provided with adjusting means A for adjusting the speed of the fan F.

[0111] In a first alternative, said adjusting means can comprise an inverter.

[0112] In a second alternative, said adjusting means can comprise one or more relay.

[0113] When said adjusting means A for adjusting the speed of the fan F comprise an inverter, the fourth group of fuzzy sets F4 comprise at least fourth fuzzy sets F4 1 ,F4 2 ,F4 3 ,F4 4 : a first fuzzy set F4 1 for identifying a first speed of the fan F, a second fuzzy set F4 2 for identifying a second speed of the fan F, different from said first speed, a third fuzzy set F4 3 for identifying a third speed of the fan F, different from said first speed and said second speed, and a fourth fuzzy set F44 for identifying a fourth speed of the fan F equal to zero.

[0114] When said adjusting means A for adjusting the speed of the fan F comprise one or more relays, said fourth group of fuzzy sets F4 comprises two respective fuzzy sets F4 1 ,F4 2 for each relay: a first fuzzy set F4 1 for identifying a first speed of the fan F equal to zero, and a secondo fuzzy set F4 2 for identifying a second speed, different from said first speed.

[0115] Regardless of whether said adjusting means comprise an inverter or one or more relays, the first real number R1 and the second real number R2 obtained through the defuzzification are both numbers greater than 0 and less than or equal to 10.

[0116] The present invention relates also to a method for controlling the thermal comfort inside an environment, through the control system above mentioned.

[0117] Said method comprises the following steps: ∘ acquiring a value of internal temperature T INi by said temperature detecting means 1 for a respective predetermined time period Δt REFi , as well as said value of temperature T U and said value of speed V U from said user interface module 2; ∘ calculating one or more values of temperature variation δT Dj , with j=1,...,M, where M is a positive integer and j is the number of predetermined time periods, and wherein each value of temperature variation δT Dj is obtained by the subtraction between the value of internal temperature T INi in a predetermined time period and the value of internal temperature T INi-1 in the previous predetermined time period; ∘ storing: the values of internal temperature T INi referred to at least two predetermined time periods Δt REFi and said one or more values of temperature variation δT Dj , as well as said temperature value T U and said value of speed V U ; a first group of fuzzy sets F1 comprising a plurality of fuzzy sets F1 1 ,F1 2 ...F1 N , each of which identifies a respective degree of belonging of a value of internal temperature T INi ; a second group of fuzzy sets F2 comprising a plurality of fuzzy sets F2 1 ,F2 2 ...F2 N , each of which identifies a respective degree of belonging of a value of temperature variation δT Dj ; a third group of fuzzy sets F3 comprising a plurality of fuzzy sets F3 1 ,F3 2 ...F3 N , each of which identifies a respective degree of belonging of an opening state of the control valve V; a fourth group of fuzzy sets F4 comprising a plurality of fuzzy sets F4 1 ,F4 2 ...F4 N , each of which identifies a respective degree of belonging of a value of the speed of the fan F; a plurality of predetermined inferential rules that establish on the basis of the degree of belonging of each value of internal temperature T INi and the degree of belonging of each value of temperature variation δT Dj , the degree of belonging of the opening state of the control valve V and the degree of belonging of the speed of the fan F, o performing a respective fuzzification referred to: each value of internal temperature T INi to obtain a respective first fuzzy data set P1 i comprising a respective plurality of first fuzzy data P1 i1 ,P1 i2 ... P1 iN , wherein each first fuzzy data identifies a respective degree of belonging of said value of internal temperature T INi , and is obtained by means of a predetermined first equation defined by a plurality of parameters; each value of temperature variation δT Dj to obtain a respective second fuzzy data set P2 i comprising a respective plurality of second fuzzy data P2 i1 ,P2 i2 ...P2 iN , wherein each second fuzzy data identifies a respective degree of belonging of said value of temperature variation δT Dj , and is obtained by means of a predetermined second equation defined by a plurality of further parameters, o applying said predetermined inferential rules to each pair of fuzzy data formed by a first fuzzy data P1 i1 ,P1 i2 ...P1 iN and a second fuzzy data P2 i1 ,P2 i2 ...P2 iN to obtain a respective third fuzzy data set P3 i comprising a plurality of third fuzzy data P3 i1 ,P3 i2 ...P3 iN , wherein each third fuzzy data identifies a respective degree of belonging of said opening state of the control valve V with respect to a respective fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3 i , and a respective fourth fuzzy data set P4 i comprising a plurality of fourth fuzzy data P4 i1 ,P4 i2 ...P4 iN , wherein each fourth fuzzy data identifies a respective degree of belonging of said value of said speed of the fan F with respect to a respective further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4 i , o performing a respective defuzzification referred to: said fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3 i to obtain a first real number R1 indicating if the control valve V is to be open and how much to open or if it is to be closed, said further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4 i to obtain a second real number R2 indicating the speed of the fan F, ∘ receiving over time a value of internal temperature T INi , a value of temperature variation δT Dj , a temperature value T U and a speed value Vu at a respective input node N IN1 ,N IN2 ,N IN3 ,N IN4 of said neural network; ∘ providing as output the updated parameters for said predetermined first equation at a respective output node of each first group of output nodes of said neural network, and the further updated parameters for said predetermined second equation at a respective output node of each second group of output nodes of said neural network; ∘ inserting said updated parameters in said predetermined first equation to obtain a first updated fuzzy data P1 i and said further updated parameters in said predetermined second equation to obtain a second updated fuzzy data P2 i , ∘ applying said predetermined inferential rules to each pair of fuzzy data formed by a first updated fuzzy data P1 i and a second updated fuzzy data P2 i , so as to obtain a third updated fuzzy data P3 i associated with the control valve V and a fourth updated fuzzy data P4 i associated with the speed of the fan F, as well as to obtain an updated first real number R1 or a second real number R2, through defuzzification; ∘ generating a signal to control said control valve V containing an information concerning said first real number R1 or said updated first real number R1 for opening / closing the control valve V and a further signal to control said adjusting means A containing an information concerning said second real number R2 or said updated second real number R2 for the speed of the fan F. Example

[0118] An example is described below in which two internal temperature values T INi are taken into consideration: a first internal temperature value T INi = 20 ° C in a predetermined first time period, a second internal temperature value T IN2 = 21,5 °C in a second predetermined period of time, subsequent to said predetermined first time period.

[0119] Each temperature value is obtained at the end of a respective predetermined time period.

[0120] Consequently, the value of temperature variation is the following: δT D1 = T IN2 -T IN1 = 1,5°C.

[0121] Furthermore, the first group of fuzzy sets F1 comprises three fuzzy sets F1 1 ,F1 2 ,F1 3 , so that three number identifying a respective degree of belonging to a fuzzy set of said first group of fuzzy sets F1 are associated with the value of internal temperature T IN1 .

[0122] In particular, the first group of fuzzy sets F1 comprises: a first fuzzy set F1 1 for identifying a first thermal condition called 'COLD', a second fuzzy set F1 2 for identifying a second thermal condition called 'GOOD', different from the first thermal condition, a third fuzzy set F1 3 for identifying a third thermal condition 'HOT', different from the first thermal condition and from the second thermal condition.

[0123] Each fuzzy set F1 1 , F1 2 , F1 3 is identified a respective first equation representing a Gaussian curve defined by a first triad of parameters a 1 , b 1 , c 1 which is predefined.

[0124] Below a first table which shows for each fuzzy set of the first group of fuzzy sets F1 a respective thermal condition and a respective first triad of parameters. Table 1Fuzzy setThermal ConditionParameters (first triad)F1 1 COLDa 1 = 0,3989 b 1 = 20 c 1 = 1,414F1 2 GOODa 1 = 0,3989 b 1 = 22 c 1 = 1,414F1 3 HOTa 1 = 0,3989 b 1 = 24 c 1 = 1,414

[0125] The second group of fuzzy sets F2 comprises three fuzzy sets F2 1 , F2 2 , F2 3 , so that three numbers identifying a respective degree of belonging to a fuzzy set of said second group of fuzzy sets F2 are associated with the value of temperature variation δT D1 .

[0126] In particular, the second group of fuzzy sets F2 comprises: a first fuzzy set F2 1 for identifying a first temperature variation called 'SLOW', a second fuzzy set F2 2 for identifying a second temperature variation called 'MEDIUM', different from the first temperature variation, a third fuzzy set F2 3 for identifying a third temperature variation call 'FAST', different from the first temperature variation and from the second temperature variation.

[0127] Each fuzzy set F2 1 , F23 2 , F2 3 is identified by a respective second equation representing a Gaussian curve defined by a second triad of parameters a 2 , b 2 , c 2 which is predefined.

[0128] Below a second table which shows for each fuzzy set of the second group of fuzzy sets F2 a respective temperature variation and a respective second triad of parameters. Table 2Fuzzy setTemperature variation temperaturaParameters (second triad)F2 1 SLOWa 2 = 0,3989 b 2 = 1 c 2 = 1,414F2 2 MEDIUMa 2 = 0,3989 b 2 = 2 c 2 = 1,414F2 3 FASTa 2 = 0,3989 b 2 = 3 c 2 = 1,414

[0129] The third group of fuzzy sets F3 comprises fourth fuzzy sets F3 1 , F3 2 , F3 3 , F3 4 : a first fuzzy set F3 1 for identifying a first open state of the control valve V: 'state 1 = OPEN1', a second fuzzy set F3 2 for identifying a second open state of the control valve V, different from the first open state: 'state 2 = OPEN2', a third fuzzy set F3 3 for identifying a third open state of the control valve V, different from the first open state and the second open state: 'state 3 = OPEN3', and a fourth fuzzy set F3 4 for identifying a closed state of the control valve V: 'state 4 = CLOSED'.

[0130] Each fuzzy set F3 1 , F3 2 , F3 3 , F3 4 is identified by a respective third equation representing a Gaussian curve defined by a third triad of parameters a 3 , b 3 , c 3 which is predefined.

[0131] Below a third table which shows for each fuzzy set of the third group of fuzzy sets F3 a respective state of the control valve and a respective third triad of parameters. Table 3Fuzzy setsState of the control valveParameters (third triad)F3 1 OPEN1a 3 = 0,3989 b 3 = 3 c 3 = 1,414F3 2 OPEN2a 3 = 0,3989 b 3 = 6 c 3 = 1,414F3 3 OPEN3a 3 = 0,3989 b 3 = 10 c 3 = 1,414F3 4 CLOSEDa 3 = 0,3989 b 3 = 0 c 3 = 1,414

[0132] In the example being disclosed, said adjusting means A for adjusting the speed of the fan F comprise an inverter.

[0133] Said fourth group of fuzzy sets F4 comprises two respective fuzzy sets F4 1 ,F4 2 for the inverter: a first fuzzy set F4 1 for identifying a first speed condition for the fan F equal to zero called 'SPEED0', and a second fuzzy set F4 2 for identifying a second speed condition for the fan F called 'SPEED1', different from the first speed condition.

[0134] Below a fourth table which shows for each fuzzy set of the fourth group of fuzzy sets F4 a respective speed condition for the fan F and a respective fourth triad of parameters. Table 4Fuzzy setsSpeed condition of the fanParameters (fourth triad)F4 1 SPEED0a 4 = 0,3989 b 4 = 2 c 4 = 1,414F4 2 SPEED1a 4 = 0,3989 b 4 = 8 c 4 = 1,414

[0135] Below a fifth table which shows the inferential rules stored in the storage means 3. Table 5Inferential rulesIf T IN1 = COLD and t D1 =SLOWV = OPEN1 and F = SPEED1If T IN1 = COLD and t D1 = MEDIUMV = OPEN2 and F = SPEED1If T IN1 = COLD and t D1 = FASTV = OPEN3 and F = SPEED1If T IN1 = GOOD and t D1 = SLOWV = CLOSED and F = SPEED0If T IN1 = GOOD and t D1 = MEDIUMV = OPEN1 and F = SPEED1If T IN1 = GOOD and t D1 = FASTV = OPEN2 and F = SPEED1If T IN1 = HOT and t D1 = SLOWV = OPEN1 and F = SPEED1If T IN1 = HOT and t D1 = MEDIUMV = OPEN2 and F = SPEED1If T IN1 = HOT and t D1 = FASTV = OPEN3 and F = SPEED1

[0136] The fuzzication referred to the value of internal temperature T IN1 = 21,5°C allows to obtain a first set of fuzzy data P1 i : P1 i1 =0,1295, P1 i2 =0,3520, P1 i3 =0,0175.

[0137] The fuzzification referred to the value of temperature variation δT D1 = 1,5°C allows to obtain a second set of fuzzy data P2 i : P2 i1 =0,3520, P2 i2 =0,3520, P2 i3 =0,1295.

[0138] The possible pairs of fuzzy data are formed by a first fuzzy data of the first fuzzy data set and by a second fuzzy data of the second fuzzy data set, wherein a thermal condition is associates with the first fuzzy data and a temperature variation is associated with the second fuzzy data.

[0139] The sixth table below shows the possible combinations of thermal condition and temperature variation. Table 6Thermal conditionTemperature variationCOLDSLOWCOLDMEDIUMCOLDFASTGOODSLOWGOODMEDIUMGOODFASTHOTSLOWHOTMEDIUMHOTFAST

[0140] The application of the inferential rules to each pair of fuzzy data formed by a thermal condition and a temperature variation (to which respective fuzzy data are associated) allows to obtain the results shown in the seventh table below (i.e. the numerical values associated with the fuzzy sets relating to the state of the control valve and the numerical values associated to the fuzzy sets relating to the fan speed). Table 7Inferential rulesThermal conditionTemperature variationState of the control valveSpeed of the fanCOLDSLOWOPEN1=0,1295SPEED1=0,1295COLDMEDIUMOPEN2=0,1295SPEED1=0,1295COLDFASTOPEN3=0,1295SPEED1=0,1295GOODSLOWCLOSED=0,3520SPEED0=0,3520GOODMEDIUMOPEN1=0,3520SPEED1 =0,3520GOODFASTOPEN2=0,1295SPEED1=0,1295HOTSLOWOPEN1=0,0175SPEED1=0,0175HOTMEDIUMOPEN2=0,0175SPEED1=0,0175HOTFASTOPEN3=0,0175SPEED1=0,0175

[0141] One or more numerical values can correspond to each fuzzy set that identifies a degree of belonging of the control valve and to each fuzzy set that identifies the fan speed.

[0142] The numerical values associated with each degree of belonging to both the control valve and the fan speed are indicated below: OPEN1 = 0,1295; 0,3520; 0,0175, OPEN2 = 0,1295; 0,1295; 0,0175, OPEN3 = 0,1295; 0; 0,0175 CLOSED = 0; 0,3520; 0, SPEEDO = 0,3520, SPEED1= 0,1295; 0,1295; 0,1295; 0,3520; 0,1295; 0,0175; 0,0175; 0,0175.

[0143] Below is the maximum value for each degree of belonging mentioned above: OPEN1 = OR(0,1295; 0,3520; 0,0175) = 0,3520, OPEN2 = OR(0,1295; 0,1295; 0,0175) = 0,1295, OPEN3 = OR(0,1295; 0; 0,0175) = 0,1295, CLOSED = OR(0; 0,3520; 0) = 0,3520, SPEED0 = 0,3520. SPEED1=OR (0,1295; 0,1295; 0,1295; 0,3520; 0,1295; 0,0175; 0,0175; 0,0175) = 0,3520.

[0144] By applying the defuzzification equations according to the method called "mean of the maxima" it is possible to obtain the following numerical values in the real domain: VALVE = 1,5V, FAN = 5V.

[0145] The temperature value T U selected by the user through the user interface module is 24°C and the speed value Vu of the fan F selected by the user is remained unchanged with respect to a value already previously selected by the user.

[0146] The neural network takes into account the temperature value T U and speed Vu selected by the user together to the internal temperature value and the temperature variation values and provides as output the triads of updated parameters shown in the following table.

[0147] Although the fan speed has not changed, the value of the fan speed is still taken into account by the neural network to obtain the updated parameters as reported in the eighth table and in the ninth table below. Table 8Fuzzy setsUpdated parameters (first triad)F1 1 =COLDa 1 = 0,3989 b 1 = 22 c 1 = 1,414F1 2 =GOODa 1 = 0,3989 b 1 = 24 c 1 = 1,414F1 3 =HOTa 1 = 0,3989 b 1 = 26 c 1 = 1,414 Table 9 Fuzzy setsUpdated parameters (second triad)F2 1 =SLOWa 2 = 0,3989 b 2 = 1 c 2 = 1,414F2 2 =NORMALa 2 = 0,3989 b 2 = 2 c 2 = 1,414F2 3 =FASTa 2 = 0,3989 b 2 = 3 c 2 = 1,414

[0148] The numerical values of the first updated parameter triad are inserted into the first equation and the numerical values of the second updated parameter triad are inserted into the second equation to obtain respectively: a respective first updated fuzzy data P1 i : P1 i1 =0,3520, P1 i2 =0,0175, P1 i3 =0; and a respective second updated fuzzy data P2 i : P2 i1 0,0175, P2 i2 =0,1295, P2 i3 =0,3520.

[0149] Therefore, the possible fuzzy data pairs are formed by a first updated fuzzy data belonging to the first fuzzy data set and by a second updated fuzzy data belonging to the second fuzzy data set, in which a thermal condition is associated with the first updated fuzzy data, and a temperature variation is associated with the second updated fuzzy datum.

[0150] As it is clear from the tenth table and from the eleventh table shown below, the thermal conditions and the temperature variations are respectively the same thermal conditions and the same temperature variations shown in table 6 but some numerical values associated with the fuzzy sets relating to the state control valve V and some numerical values associated with the fuzzy sets related to the fan speed F have changed. Table 10Thermal conditionTemperature variationCOLDSLOWCOLDMEDIUMCOLDFASTGOODSLOWGOODMEDIUMGOODFASTHOTSLOWHOTMEDIUMHOTFAST

[0151] The application of the inferential rules to each pair of data formed by an updated thermal condition and a temperature variation (to which respective updated fuzzy data are associated) allows to obtain the results shown in the eleventh table below. Table 11Inferential rulesThermal conditionTemperature variationState of the control valveSpeed of the fanCOLDSLOWOPEN1=0,0175SPEED1=0,0175COLDMEDIUMOPEN2=0,1295SPEED1=0,1295COLDFASTOPEN3=0,3520SPEED1 =0,3520GOODSLOWCLOSED=0,0175SPEED0=0,0175GOODMEDIUMOPEN1=0,0175SPEED1=0,0175GOODFASTOPEN2=0,0175SPEED1=0,0175HOTSLOWOPEN1=0SPEED1=0HOTMEDIUMOPEN2=0SPEED1=0HOTFASTOPEN3=0SPEED1=0

[0152] By comparing table 11 with table 7, the numerical values of some third updated fuzzy data referring to the status of the control valve V and the numerical values of some fourth updated fuzzy data referring to the fan speed have changed.

[0153] One or more updated numerical values can correspond to each fuzzy set that identifies a degree of belonging of the control valve and to each fuzzy set that identifies a degree of belonging of the fan speed.

[0154] The updated numerical values associated with each degree of belonging to both the control valve and the fan speed are shown below: OPEN1 = 0,0175; 0,0175; 0, OPEN2 = 0,1295; 0,0175; 0, OPEN3 = 0,3520; 0; 0, CLOSED = 0; 0,0175; 0, SPEED0 = 0,0175, SPEED1 = 0,0175; 0,1295;0,3520; 0,0175; 0,0175; 0; 0; 0.

[0155] Below the maximum value for each degree of belonging mentioned above is shown: OPEN1 = OR(0,0175; 0,0175; 0) = 0,0175, OPEN2 = OR(0,1295; 0,0175; 0) = 0,1295, OPEN3 = OR(0,3520; 0; 0) = 0,3520, CLOSED = OR(0; 0,0175; 0) = 0,0175, SPEED0 = 0,0175, SPEED1 = OR(0,0175; 0,1295;0,3520; 0,0175; 0,0175; 0; 0; 0) = 0,3520.

[0156] By applying the defuzzification equations according to the method called "mean of the maxima", the following numerical values are obtained in the real domain: VALVE = 10V, FAN = 8V.

[0157] The numerical value associated with the state of the control valve V and the numerical value associated with the fan speed F have changed with respect to the numerical values mentioned above (i.e. 1.5V for the valve and 5V for the fan).

[0158] Each numerical value associated with the state of the control valve V and each numerical value associated with the fan speed F are expressed in Volts.

[0159] Finally, the control logic unit 4 will generate a signal to control the control valve V and a further signal to control said adjusting means A of said air conditioning unit U (taking into account the information obtained by applying the inferential rules to the updated fuzzy data pairs) in order to check the thermal comfort of an environment in which a user is present, based on the user's needs.

[0160] In particular, said signal and said further signal are respective voltage signals.Advantages

[0161] Advantageously, through the control system object of the invention, it is possible to control the thermal comfort level inside an environment taking into account of the needs of a user inside said environment.

[0162] In particular, the use of a neuro-fuzzy controller using a fuzzy logic in combination with a neural network allows an adaptive control of the thermal comfort level.

[0163] A second advantage is given by the fact that the control system is adapted to control the thermal comfort reducing the energy consumption by means of an appropriate use of the resources available to said user.

[0164] A third advantage is given by the fact that the control system object of the invention can be easily used both by a user and by an operator assigned to the maintenance and management of said control system.

[0165] A fourth advantage is to offer the user the possibility of actively participating in the definition of the thermal comfort conditions of the control system, modifying the values of parameters useful for determining the thermal comfort.

[0166] A further advantage is given by the possibility of reducing the differences between what is described by the rules on well-being and the real perception of well-being by a user, without distorting the indications imposed by these rules.

[0167] The present invention has been described for illustrative, but not limitative purposes, according to its preferred embodiment, but it is to be understood that variations and / or modifications can be carried out by a skilled in the art, without departing from the scope thereof, as defined according to enclosed claims.

Claims

1. Control system adapted for controlling at least one air conditioning unit (U) positioned in an environment and through which the thermal comfort of said environment is controlled, said air conditioning unit (U) being of the type comprising a fan (F), adjusting means (A) for adjusting the speed of said fan (F), a control valve (V) for controlling the flow rate of a thermovector fluid over time, said control system comprising: - temperature detecting means (1), to be installed in said environment, configured to detect a respective value of internal temperature TINi of said environment for a plurality of predetermined time periods ΔtREFi, with i=1,...,N, wherein N is a positive integer, - a user interface module (2) configured to allow a user to select a temperature value Tu and a speed value Vu of the fan (V), - storage means (3), - a logic control unit (4) configured to: ∘ acquire a value of internal temperature TINi from said temperature detecting means (1) for a respective predetermined time period ΔtREFi, as well as said temperature value TU and said speed value VU from said user interface module (2), ∘ calculate one or more values of temperature variations δTDj, with j=1,...,M, wherein M is a positive integer and j is the number of predetermined time periods, and wherein each value of temperature variation δTDj is obtained by the subtraction between the value of internal temperature TINi in a predetermined time period and the value of internal temperature TINi-1 in the previous predetermined time period; ∘ store in said storage means (3): the values of internal temperature TINi referred to at least two predetermined time periods ΔtREFi and said one or more values of temperature variation δTDj, as well as said temperature value TU and said speed value VU, a first group of fuzzy sets F1 comprising a plurality of fuzzy sets F11,F12...F1N, each of which identifies a respective degree of belonging of a value of internal temperature TINi, a second group of fuzzy sets F2 comprising a plurality of fuzzy sets F21,F22...F2N, each of which identifies a respective degree of belonging of a value of temperature variation δTDj, a third group of fuzzy sets F3 comprising a plurality of fuzzy sets F31,F32...F3N, each of which identifies a respective degree of belonging of an opening state of the control valve (V), a fourth group of fuzzy sets F4 comprising a plurality of fuzzy sets F41,F42...F4N, each of which identifies a respective degree of belonging of a value of the speed of the fan (F), a plurality of predetermined inferential rules that establish on the basis of the degree of belonging of each value of internal temperature TINi and the degree of belonging of each value of temperature variation δTDj, the degree of belonging of the opening state of the control valve (V) and the degree of belonging of the speed of the fan (F), o perform a respective fuzzification referred to: each value of internal temperature TINi to obtain a respective first fuzzy data set P1i comprising a respective plurality of first fuzzy data P1i1,P1 i2... P1iN, wherein each first fuzzy data identifies a respective degree of belonging of said value of internal temperature TINí, and is obtained by means of a predetermined first equation defined by a plurality of parameters, each value of temperature variation δTDj to obtain a respective second fuzzy data set P2i comprising a respective plurality of second fuzzy data P2i1,P2i2...P2iN, wherein each second fuzzy data identifies a respective degree of belonging of said value of temperature variation δTDj, and is obtained by means of a predetermined second equation defined by a plurality of further parameters, ∘ apply said predetermined inferential rules to each pair of fuzzy data formed by a first fuzzy data P1i1,P1i2...P1iN and a second fuzzy data P2i1,P2i2...P2iN to obtain a respective third fuzzy data set P3i comprising a plurality of third fuzzy data P3i1,P3i2...P3iN, wherein each third fuzzy data identifies a respective degree of belonging of said opening state of the control valve (V) with a respective fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3i, and a respective fourth fuzzy data set P4i comprising a plurality of fourth fuzzy data P4i1,P4i2...P4iN, wherein each fourth fuzzy data identifies a respective degree of belonging of said value of the speed of the fan (F) with respect to a respective further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4i, ∘ perform a respective defuzzification referred to: said fuzzy data set associated with to said third fuzzy data of said third fuzzy data set P3i to obtain a first real number R1 indicating if the control valve (V) is to be open and how much to open or if it is to be closed, said further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4i to obtain a second real number R2 indicating the speed of the fan (F), ∘ perform a neural network comprising: an input layer comprising a plurality of input nodes (NIN1,NIN2,NIN3,NIN4), and an output layer comprising a respective first group of output nodes (NOUT1F11, NOUT2F11...NOUTNF11; NOUT1F12,NOUT2F12...NOUTNF12;... NOUT1F1N,NOUT2F1N...NOUTNF1N) for each fuzzy set F11,F12...F1N of the first group of fuzzy sets F1, as well as a respective second group of output nodes (NOUT1F21,NOUT2F21...NOUTNF21; NOUT1F22,NOUT2F22...NOUTNF22;... NOUT1F2N, NOUT2F2N...NOUTNF2N) for each fuzzy set F21,F22...F2N of the second group of fuzzy sets F2, wherein said neural network is configured to: • receive over time said value of internal temperature TINi, said value of temperature variation δTDj, said temperature value Tu and said speed value Vu at a respective input layer (NIN1,NIN2,NIN3,NIN4), and • provide as output the updated parameters for said predetermined first equation at a respective output node of each first group of output nodes, and the further updated parameters for said predetermined second equation at a respective output node of each second group of output nodes, ∘ insert said updated parameters in said predetermined first equation to obtain a first updated fuzzy data P1i and said further updated parameters in said predetermined second equation to obtain a second updated fuzzy data P2i, ∘ apply said predetermined inferential rules to each pair of fuzzy data formed by a first updated fuzzy data P1i and a second updated fuzzy data P2i, so as to obtain a third updated fuzzy data P3i associated with the control valve (V) and a fourth updated fuzzy data P4i associated with the speed of the fan (F), as well as to obtain an updated first real number R1 and an updated second real number R2, through defuzzification; ∘ generate a signal to control said control valve (V) containing an information concerning said first real number R1 or said updated first real number R1 for opening / closing the control valve (V) and a further signal to control said adjusting means (A) containing an information concerning said second real number R2 or said updated second real number R2 for the speed of the fan (F).

2. Control system according to the previous claim, wherein with reference to the fuzzification referred to each value of internal temperature TINi, said predetermined first equation is defined by a first triad of parameters a1, b1, c1 and represents a Gaussian curve: P 1 i = a 1 ⋅ e − x i − b 1 2 c 1 2 where the parameter a1 is a predetermined number greater than 0 and less than or equal to 1 and represents the height of the Gaussian curve; the parameter b1 is a predetermined value of temperature in which the Gaussian curve is centred; the parameter c1 is a predetermined value indicating the distance between a symmetry axis of the Gaussian curve and an inflection point of the Gaussian curve; and xi is a respective value of internal temperature TINi, wherein with reference to the fuzzification referred to each value of temperature variation δTDj, said predetermined second equation is defined by a second triad of parameters a2, b2, c2 and represents a Gaussian curve: P 2 i = a 2 ⋅ e − x j − b 2 2 c 2 2 where the parameter a2 is a predetermined number greater than 0 and less than or equal to 1 and represents the height of the Gaussian curve; the parameter b2 is a predetermined value of temperature in which the Gaussian curve is centred; the parameter c2 is a predetermined value indicating the distance between a symmetry axis of the Gaussian curve and an inflection point of the Gaussian curve; and xj is a respective value of temperature variation δTDj.

3. Control system according to the claim 2, wherein each first group of output nodes (NOUT1F11, NOUT2F11...NOUTNF11;NOUT1F12,NOUT2F12...NOUTNF12;...NOUT1F1N,NOUT2F1N...NOUT NF1N) comprising three respective output nodes (NOUT1F11, NOUT2F11,NOUT3F11; NOUT1F12,NOUT2F12,NOUT3F12;...NOUT1Fl1N,NOUT2F1N,NOUT3F1N) and said neural network is configured to provide at each output node of said first group of output nodes a respective updated parameter a1, b1, c1, so as to obtain a respective first triad of updated parameters a1, b1, c1, and said logic control unit (4) is configured to insert said first triad of updated parameters a1, b1, c1 in said predetermined first equation to obtain a first updated fuzzy data P1i; wherein each second group of output nodes (NOUT1F21,NOUT2F21...NOUTNF21; NOUT1F22,NOUT2F22...NOUTNF22;.. NOUT1F2N, NOUT2F2N...NOUTNF2N) comprising three respective output nodes (NOUT1F21,NOUT2F21,NOUT3F21;NOUT1F21,NOUT2F21,NOUT3F21;...NOUT1F2N,NOUT2F2N, NOUT3F2N) and said neural network is configured to provide at each output node of said second group of output nodes a respective updated parameter a2, b2, c2, so as to obtain a second triad of updated parameters a2, b2, c2, and said logic control unit (4) is configured to insert said second triad of updated parameters a2, b2, c2 in said predetermined second equation to obtain a second updated fuzzy data P2i.

4. Control system according to claim 2 or 3, wherein the defuzzification referred to said fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3i is performed trough a method called mean of maxima and the logic control unit (4) is configured to: ∘ select a maximum value P3iMAX with respect to each third fuzzy data P3i1,P3i2...P3iN, wherein said maximum value is associated with at least one fuzzy set F31,F32...F3N of the third group of fuzzy sets F3, ∘ calculate at least two real numbers P31, P3i applying a predetermined third equation defined by a third triad of parameters a3, b3, c3: x 2 − 2 b 3 x + b 3 2 + c 3 2 ln P 3 iMAX a 3 = 0 wherein the parameter a3 is a number greater than 0 and less than or equal to 1 and represents the height of a Gaussian curve; the parameter b3 is a value of temperature in which a Gaussian curve is centred; the parameter c3 is the value of the distance between a symmetry axis of a Gaussian curve and an inflection point of said Gaussian curve; and x is a value that P3 can assume; ∘ calculate said first real number R1 equal to the mean value of said at least two real numbers P31, P32; and wherein the defuzzification referred to said further fuzzy data associated with said fourth fuzzy data of said fourth fuzzy data set P4i is performed through a method called mean of maxima and the logic control unit (4) is configured to: ∘ select a maximum value P4iMAX with respect to each fourth fuzzy data P4i1,P4i2...P4iN, wherein said maximum value is associated with at least one fuzzy set F41,F42...F4N of the fourth group of fuzzy sets F4, ∘ calculate at least two reals number P41, P42 applying the following predetermined fourth equation defined by a forth triad of parameters a4, b4, c4: x 2 − 2 b 4 x + b 4 2 + c 4 2 ln P 4 iMAX a 4 = 0 where the parameter a4 is a number greater than 0 and less than or equal to 1 and represents the height of a Gaussian curve; the parameter b4 is a value of temperature in which a Gaussian curve is centred; the parameter c4 is the value of the distance between a symmetry axis of a Gaussian curve and an inflection point of said Gaussian curve; and x is the value that P4 can assume; calculate said second real number R2 equal to the mean value of said at least two real numbers P41, P42.

5. Control system according to the previous claim, wherein the output layer of said neural network comprises: ▪ a respective third group of output nodes (NOUT1F31, NOUT2F31...NOUTNF31;NOUT1F32,NOUT2F32...NOUTNF32;... NOUT1F3N,NOUT2F3N...NOUTNF3N) for each fuzzy set F31,F32...F3N of the third group of fuzzy sets F3, ▪ a respective fourth group of output nodes (NOUT1F41,NOUT2F41...NOUTNF41;NOUT1F42,NOUT2F42...NOUTNF42;...NOUT1F4N ,NOUT2F4N...NOUTNF4N) for each fuzzy set F41,F42...F4N of the fourth group of fuzzy sets F4, wherein said neural network is configured to provide as output: the updated parameters a3, b3, c3 of said third triad of parameters for said predetermined third equation, each of which is at a respective output node of each third group of output nodes, the updated parameters a4, b4, c4 of said fourth triad of parameters for said predetermined fourth equation, each of which is at an output node of each fourth group of output nodes, wherein said logic control unit (4) is configured to: ∘ insert said third triad of updated parameters a3, b3, c3 in said predetermined third equation to obtain an updated fuzzy data set, due to the change of the shape of the function represented by said predetermined third equation, wherein said updated fuzzy data set is associated with said third fuzzy data of said third fuzzy data set P3i, o insert said fourth triad of updated parameters a4, b4, c4 in said predetermined fourth equation to obtain an updated further fuzzy data set, due to the change of the shape of the function represented by said predetermined fourth equation, wherein said updated further fuzzy data set is associated with said fourth fuzzy data of said fourth fuzzy data set P4i, defuzzify each updated fuzzy data set and each updated further fuzzy data set.

6. Control system according to the previous claim, wherein each third group of output nodes (NOUT1F31, NOUT2F31...NOUTNF31;NOUT1F32,NOUT2F32...NOUTNF32;... NOUT1F3N,NOUT2F3N...NOUTNF3N) comprises three output nodes (NOUT1F31, NOUT2F31,NOUT3F31;NOUT1F32,NOUT2F32,NOUT3F32;...NOUT1F3N,NOUT2F3N,NOUT3F3N) and said neural network is configured to provide a respective updated parameter a3, b3, c3 for said third equation at each output node of said third group of output nodes, so as to obtain a respective third triad of updated parameters a3, b3, c3 for each third group of output nodes, and said logic control unit (4) is configured to insert said updated parameters a3, b3, c3 in said predetermined third equation to obtain an updated fuzzy data set, due to the change of the shape of the function represented by said predetermined third equation, wherein said updated fuzzy data set is associated with said third fuzzy data of said third fuzzy data set P3i, wherein each fourth group of output nodes (NOUT1F41, NOUT2F41...NOUTNF41;NOUT1F42,NOUT2F42...NOUTNF42;...NOUT1F4N,NOUT2F4N...NOUT NF4N) comprises three respective output nodes (NOUT1F41,NOUT2F41,NOUT3F41;NOUT1F42,NOUT2F42...NOUTNF42;...NOUT1F4N,NOUT2F4 N...NOUTNF4N), in which a respective updated parameter a4, b4, c4 is present, so as to obtain a respective fourth triad of updated parameters a4, b4, c4 for each fourth group of output nodes, and said logic control unit (4) is configured to insert said updated parameters a4, b4, c4 in said predetermined fourth equation to obtain an updated further fuzzy data set, due to the change of the shape of the function represented by said predetermined fourth equation, wherein said updated further fuzzy data set is associated with said fourth fuzzy data of said fourth fuzzy data set P4i.

7. Control system according any one of the previous claims, wherein said neural network is Kohonen network and said input layer (NIN2) for receiving said value of temperature Tu selected by the user and said input layer (NIN3) for receiving said value of speed Vu of the fan (F) selected by the user are arranged between said input node (NIN1) for receiving a value of internal temperature TINi and said input node (NIN4) for receiving a value of temperature variation δTDj.

8. Control system according to the claim 1, wherein said predetermined first equation for the fuzzification of each value of internal temperature TINi is any one function identifying a degree of belonging to a fuzzy set of the first group of fuzzy sets, and said predetermined second equation for the fuzzification of each value of temperature variation δTDj is any one function identifying a degree of belonging to a fuzzy set of the second group of fuzzy sets.

9. Control system according to any one of claims 1-3, wherein said defuzzification is performed by means of one of the following methods: last of maxima, center of sums, center of area, center of inertia, center of gravity.

10. Control System according to any one of the previous claims, wherein Said first group of fuzzy sets F1 comprises at least three fuzzy sets F11,F12,F13, so that at least three numbers identifying a respective degree of belonging to a fuzzy set of said first group of fuzzy sets F1 are associated with to each value of internal temperature TINi, wherein said second group of fuzzy sets F2 comprises at least three fuzzy sets F21,F22,F23, so that at least three numbers identifying a respective degree of belonging to a fuzzy set of said second group of fuzzy sets F2 are associated to each value of temperature variation δTDj.

11. Control system according to any one of the previous claims, wherein said third group of fuzzy sets F3 comprises at least four fuzzy sets F31,F32,F33,F34: a first fuzzy set F31 to identify a first opening state of the control valve (4), a second fuzzy set F32 to identify a second opening state of the control valve (V), different from said first opening state, a third set fuzzy F33 to identify a third opening state of the control valve (V), different from said first opening state and said second opening state, and a fourth fuzzy set F34 to identify a closing state of the control valve (V).

12. Control system according to any one of claims 1-11, wherein, when said adjusting means (A) to adjust the speed of the fan (F) comprise an inverter, said fourth group of fuzzy sets F4 comprises at least four fuzzy sets F41,F42,F43,F44: a first fuzzy set F41 to identify a first speed of the fan (F), a second fuzzy set F42 to identify a second speed of the fan (F), different from said first speed, a third fuzzy set F43 to identify a third speed of the fan (F), different from said first speed and from said second speed, and a fourth fuzzy set F44 to identify a fourth speed of the fan (F) equal to zero; said first real number R1 being preferable greater than 0 and less than or equal to 10, and said second real number R2 being preferable greater than 0 and less than 10.

13. Control system according to any one of claims 1-11, wherein, when said adjusting means (A) to adjust the speed of the fan (F) comprises one or more relay, said fourth group of fuzzy sets F4 comprises two respective fuzzy sets F41,F42 for each relay: a first fuzzy set F41 to identify a first speed of the fan (F) equal to zero and a second fuzzy set F42 to identify a second speed, different from said first speed; said first real number R1 being preferable greater than 0 and less than or equal to 10, and said second real number R2 being preferable greater than 0 and less than 10.

14. Air conditioning unit (U) comprising a control system according to any one of the previous claims.

15. Method for controlling, through the control system according to any one of the claims 1-14, at least one air conditioning unit (U) positioned in an environment and through which the thermal comfort of said environment is controlled, said air conditioning unit (U) being of the type comprising a fan (F), adjusting means (A) for adjusting the speed of said fan (F), a control valve (V) for controlling the flow rate of a thermovector fluid over time, said method comprising the following steps: ∘ acquiring a value of internal temperature TINi for a respective predetermined time period ΔtREFi, as well as said value of temperature TU and said value of speed VU from said user interface module (2); ∘ calculating one or more values of temperature variation δTDj, with j=1,...,M, where M is a positive integer and j is the number of predetermined time periods, and wherein each value of temperature variation δTDj is obtained by the subtraction between the value of internal temperature TINi in a predetermined time period and the value of internal temperature TINi-1 in the previous predetermined time period; ∘ storing: the values of internal temperature TINi referred to at least two predetermined time periods ΔREFi and said one or more values of temperature variation δTDj, as well as said temperature value TU and said value of speed VU; a first group of fuzzy sets F1 comprising a plurality of fuzzy sets F11,F12...F1N, each of which identifies a respective degree of belonging of a value of internal temperature TINi; a second group of fuzzy sets F2 comprising a plurality of fuzzy sets F21,F22...F2N, each of which identifies a respective degree of belonging of a value of temperature variation δTDj; a third group of fuzzy sets F3 comprising a plurality of fuzzy sets F31,F32...F3N, each of which identifies a respective degree of belonging of an opening state of the control valve (V); a fourth group of fuzzy sets F4 comprising a plurality of fuzzy sets F41,F42...F4N, each of which identifies a respective degree of belonging of a value of the speed of the fan (F); a plurality of predetermined inferential rules that establish on the basis of the degree of belonging of each value of internal temperature TINi and the degree of belonging of each value of temperature variation δTDj, the degree of belonging of the opening state of the control valve (V) and the degree of belonging of the speed of the fan (F), ∘ performing a respective fuzzification referred to: each value of internal temperature TINi to obtain a respective first fuzzy data set P1i comprising a respective plurality of first fuzzy data P1i1,P1i2...P1iN, wherein each first fuzzy data identifies a respective degree of belonging of said value of internal temperature TINí, and is obtained by means of a predetermined first equation defined by a plurality of parameters; each value of temperature variation δTDj to obtain a respective second fuzzy data set P2i comprising a respective plurality of second fuzzy data P2i1,P2i2...P2iN, wherein each second fuzzy data identifies a respective degree of belonging of said value of temperature variation δTDj, and is obtained by means of a predetermined second equation defined by a plurality of further parameters, ∘ applying said predetermined inferential rules to each pair of fuzzy data formed by a first fuzzy data P1i1,P1 12...P1iN and a second fuzzy data P2i1,P2i2...P2iN to obtain a respective third fuzzy data set P3i comprising a plurality of third fuzzy data P3i1,P3i2...P3iN, wherein each third fuzzy data identifies a respective degree of belonging of said opening state of the control valve (V) with respect to a respective fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3i, and a respective fourth fuzzy data set P4i comprising a plurality of fourth fuzzy data P4i1,P4i2...P4iN, wherein each fourth fuzzy data identifies a respective degree of belonging of said value of said speed of the fan (F) with respect to a respective further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4i, ∘ performing a respective defuzzification referred to: said fuzzy data set associated with said third fuzzy data of said third fuzzy data set P3i to obtain a first real number R1 indicating if the control valve (V) is to be open and how much to open or if it is to be closed, said further fuzzy data set associated with said fourth fuzzy data of said fourth fuzzy data set P4i to obtain a second real number R2 indicating the speed of the fan (F), o receiving over time a value of internal temperature TINi, a value of temperature variation δTDj, a temperature value Tu and a speed value Vu at a respective input node (NIN1,NIN2,NIN3,NIN4) of said neural network; o providing as output the updated parameters for said predetermined first equation at a respective output node of each first group of output nodes of said neural network, and the further updated parameters for said predetermined second equation at a respective output node of each second group of output nodes of said neural network; o inserting said updated parameters in said predetermined first equation to obtain a first updated fuzzy data P1i and said further updated parameters in said predetermined second equation to obtain a second updated fuzzy data P2i, o applying said predetermined inferential rules to each pair of fuzzy data formed by a first updated fuzzy data P1i and a second updated fuzzy data P2i, so as to obtain a third updated fuzzy data P3i associated with the control valve (V) and a fourth updated fuzzy data P4i associated with the speed of the fan (F), as well as to obtain an updated first real number R1 or a second real number R2, through defuzzification; o generating a signal to control said control valve (V) containing an information concerning said first real number R1 or said updated first real number R1 for opening / closing the control valve (V) and a further signal to control said adjusting means (A) containing an information concerning said second real number R2 or said updated second real number R2 for the speed of the fan (F).