Machine to dispense coffee-based beverages, and corresponding dispensing method and program

1. Machine to dispense liquid coffee comprising a circuit provided with at least: a pump configured to feed a controlled quantity of water under pressure, a heating device configured to heat the water supplied by said pump, an extraction chamber located downstream of said heating device and configured to contain a desired quantity of coffee powder, and a dispensing valve of the proportional type and selectively adjustable to control the dispensing flow of said liquid coffee exiting from said extraction chamber, and wherein said machine comprises sensors configured to detect at least one functioning parameter of said circuit, a user interface with which a user can select one of a plurality of liquid coffee recipes, and a memorization device to memorize a list of characteristic extraction curves of the liquid coffee, each associated with one of said recipes, wherein said sensors are configured to detect, repeatedly during the dispensing time, said at least one functioning parameter of said circuit, and wherein said machine comprises a control and command unit configured to implement:

a non-linear mathematical model with coupled variables, configured to mathematically describe the functioning of said circuit,

a processing and filtering algorithm configured to receive said at least one functioning parameter detected by said sensor, and to determine, repeatedly during the dispensing time and on the basis of said mathematical model, a plurality of system variables suitable to describe the behavior of said circuit over time,

a non-linear control algorithm configured to compare and identify, repeatedly during the dispensing time, a deviation between said system variables and said characteristic extraction curve selected and to supply, on exit, repeatedly during the dispensing time, command signals at least of said pump, said heating device and said dispensing valve.

2. Machine as in claim 1 , wherein said mathematical model is configured to implement neural networks suitable to process said functioning parameters detected and to determine said system variables.

3. Machine as in claim 1 , wherein said mathematical model comprises a plurality of non-linear first principle equations that govern the physical phenomena of the circuit and that are used to determine said system variables.

4. Machine as in claim 1 , wherein said mathematical model comprises, or implements, a plurality of correlation maps configured to correlate said system variables with respect to each other.

5. Machine as in claim 1 , wherein said processing and filtering algorithm comprises one or more filters configured to filter the disturbances and instabilities of the functioning parameters detected by the sensors.

6. Machine as in claim 1 , wherein the processing and filtering algorithm comprises an “Unscented Kalman Filter”.

7. Machine as in claim 1 , wherein said non-linear control algorithm can comprise a model predictive control of the adaptive type.

8. Machine as in claim 1 , wherein said pump is of the micro-gear type.

9. Machine as in claim 1 , wherein said heating device is the rapid type, and is configured to have a time to heat the water, in order to reach a desired temperature, of less than 3 seconds.

10. Machine as in claim 9 , wherein said heating device is configured to have a time to heat the water, in order to reach a desired temperature, of less than 2 seconds.

11. Machine as in claim 1 , wherein said dispensing valve is a valve functioning with a pulse width modulation logic (PWM).

12. Machine as in claim 1 , wherein said dispensing valve is an electromagnetically driven valve.

13. Method to dispense liquid coffee through a circuit, comprising:

feeding a controlled quantity of water under pressure by means of a pump;

heating the water fed from said pump with a heating device;

positioning a desired quantity of coffee powder in an extraction chamber;

making the heated water pass from said pump through said extraction chambers;

adjusting the dispensing flow of said liquid coffee exiting from said extraction chamber with a dispensing valve of the proportional type and selectively adjustable;

detecting with sensors at least one functioning parameter of said circuit;

selecting one of a plurality of recipes for liquid coffee by means of a user interface, to memorize a list of characteristic extraction curves of the liquid coffee, each associated with one of said recipes, in a memorization device, wherein said sensors detect, repeatedly during the dispensing time, said at least one functioning parameter of said circuit, and wherein said method further comprising:

receiving said at least one functioning parameter and determining with a processing and filtering algorithm, repeatedly during the dispensing time, a plurality of system variables suitable to describe the behavior of said circuit over time, said system variables also being determined on the basis at least of a non-linear mathematical model with coupled variables, configured to mathematically describe the functioning of said circuit,

comparing and identifying, repeatedly during the dispensing time, with a non-linear control algorithm, a deviation between said system variables and said characteristic extraction curve selected,

supplying, on exit, repeatedly during the dispensing time, command signals at least of said pump, said heating device and said dispensing valve.

14. Method as in claim 13 , wherein said command signals are supplied at entrance to said mathematical model.

15. Method as in claim 12 , wherein said functioning parameter comprises the flow rate of water supplied by the pump and wherein said flow rate is detected by said sensors with a sampling frequency comprised between 20Hz and 50Hz.

16. Method as in claim 12 , wherein said functioning parameter comprises the temperature of the water in transit in the circuit, and wherein said flow rate is detected by said sensors with a sampling frequency comprised between 5Hz and 10Hz.