SYSTEM FOR MANAGING THE THERMAL COMFORT OF A PERSON

Abstract

The present invention relates to a system (1) for managing the thermal comfort of a person, in particular on board a motor vehicle, this system being designed to use at least one state characteristic of the person and/or of their thermal environment, this characteristic being able to adopt a plurality of values, this system being designed to: acquire, using a sensor, data relating to the person and/or to their thermal environment, this sensor being in particular a camera designed to acquire an image of the person and/or of their thermal environment, evaluate, on the basis of these acquired data, the probability of the state characteristic adopting a first value by associating a first confidence level with this first value.

Claims

1. A system for managing the thermal comfort of a person, in particular on board a motor vehicle, the system being configured to use at least one state characteristic of the person and/or of their thermal environment, this characteristic being able to adopt a plurality of values, the system comprising: A sensor for acquiring data relating to the person and/or to their thermal environment, the sensor comprising a camera configured to acquire an image of the person and/or of their thermal environment, wherein the system is configured to: evaluate, on the basis of the acquired data, the probability of the state characteristic adopting a first value by associating a first confidence level with this first value; evaluate, on the basis of the acquired data, the probability of the state characteristic adopting a second value by associating a second confidence level with this first value, assign, to the state characteristic, an adjusted value that is a function of at least said first and second values and said associated confidence levels.

2. The system as claimed in claim 1, wherein the state characteristic, is a characteristic that may adopt multiple values, each value corresponding to a category chosen from among a plurality of categories.

3. The system as claimed in claim 1, wherein the state characteristic is a gender of the passenger that may adopt a value corresponding to a Male category and a value corresponding to a Female category.

4. The system as claimed in claim 1, wherein the system is configured to determine a confidence level for each value of the state characteristic, in particular using one or more neural networks of the system, or a Bayesian network of the system, or any other sensor and processing algorithm.

5. The system as claimed in claim 1, wherein the system comprises at least one measurement chain comprising a sensor, an acquisition unit and a processing unit running an algorithm, this measurement chain being designed to provide, for the state characteristic, values, values corresponding in particular to categories, and confidence levels associated with these values.

6. The system as claimed in claim 1, wherein the system is configured to assign the measurement chain a confidence level specific to the measurement chain, this confidence level reflecting the reliability of this measurement chain to give a reliable result.

7. The system as claimed in claim 1, further comprising: a plurality of measurement chains for determining values of the state characteristic, each measurement chain being configured to give values for the state characteristic with an associated confidence level (Prob(Aij)) such that the system assigns each value (Ai) an overall confidence level (Prob(Ai)) that is a function of the confidence levels (Conf(Cj)) specific to each measurement chain and of the confidence levels (Prob(Aij)) associated with each value (Aij) given by the measurement chain.

8. The system as claimed in claim 7, wherein the system is configured to assign the state characteristic a continuous value (E*), which is a function of values g(Ai) adopted for each value or category (Ai) of another state characteristic (A), weighted by the overall confidence level Prob(Ai) associated with this value or category (Ai), wherein E*=f([Σi(Prob(Ai)*g(Ai))/Σi(Prob(Ai))], Bk), where f is a model or algorithm and Bk is one or more other characteristics.

9. The system as claimed in claim 1, wherein the system is configured to assign the state characteristic (E) a value (E*) that is a function of values (Ei) adopted by E for each value or category (Ai) of another state characteristic (A), weighted by the overall confidence level Prob(Ai) associated with this value or category (Ai), wherein E*=Σi(Prob(Ai)*f(g(Ai), Bk))/Σi(Prob(Ai)), where f is a model or algorithm and Bk is one or more other characteristics, and Ei=f(g(Ai), Bk), g(Ai) being in particular a discrete value.

10. A method for managing the thermal comfort of a person, on board a motor vehicle, the method using at least one state characteristic of the person and/or of their thermal environment, the state characteristic being able to adopt a plurality of values, the method comprising: acquiring, using a sensor, data relating to the person and/or to their thermal environment, the sensor being in particular a camera configured to acquire an image of the person and/or of their thermal environment; evaluating, on the basis of these acquired data, the probability of the state characteristic adopting a first value by associating a first confidence level with this first value; evaluating, on the basis of these acquired data, the probability of the state characteristic adopting a second value by associating a second confidence level with this first value; and assigning, to the state characteristic, an adjusted value that is a function of at least said first and second values and said associated confidence levels.

Description

[0045] The invention and its various applications will be better understood upon reading the following description and with reference to the accompanying figures, in which:

[0046] FIG. 1 schematically illustrates a thermal comfort management system according to one non-limiting embodiment of the invention.

[0047] FIG. 1 shows a system 1 for managing the thermal comfort of a person on board a motor vehicle, this system comprising multiple sensors including: [0048] a DMS (Driver Monitoring System) camera 3 designed to observe a passenger in the passenger compartment, [0049] one or more domes 4 comprising one or more cameras operating in the visible (RGB) and/or near infrared (NIR) and/or far infrared (FIR) spectrum, placed on a ceiling of the passenger compartment and making it possible to view walls of the passenger compartment and certain body parts of the passengers, [0050] an outdoor temperature sensor and an insolation sensor 5, [0051] at least one air temperature sensor 6 at the outlet of an air conditioning device or of the HVAC 10, [0052] at least one sensor for detecting the air flows and their distribution at the outlet of an air conditioning device or of the HVAC 10, [0053] at least one sensor 7 for sensing the air temperature prevailing in the passenger compartment, [0054] preferably a humidity sensor and temperature sensors arranged in certain walls of the passenger compartment, [0055] preferably a sensor for sensing the heat flow in areas in contact with the passengers.

[0056] The system 1 is designed to use at least one state characteristic of the person and/or of their thermal environment, this characteristic being able to adopt a plurality of values, these values being discrete or continuous, this system comprising a processing unit 2 designed to: [0057] a. acquire, using one of the sensors mentioned above, data relating to the person and/or to their thermal environment, this sensor being in particular a camera designed to acquire an image of the person and/or of their thermal environment, [0058] b. evaluate, on the basis of these acquired data, the probability of the state characteristic adopting a first value by associating a first confidence level with this first value, [0059] C. evaluate, on the basis of these acquired data, the probability of the state characteristic adopting a second value by associating a second confidence level with this first value, [0060] d. assign, to the state characteristic, an adjusted value that is a function of at least said first and second values and said associated confidence levels.

[0061] The system 1 is designed to assign the state characteristic (E) a value, in particular a continuous value, which is a function of values g(Ai) adopted for each value or category (Ai) of another state characteristic (A), weighted by the overall confidence level Prob(Ai) associated with this value or category (Ai). For example, it is possible to have the formula E=f([Σi(Prob(Ai)*g(Ai))/Σi (Prob(Ai))], Bk . . . ), where f is a model or algorithm and Bk is one or more other characteristics.

[0062] According to one of the aspects of the invention, the system 1 is designed to assign the state characteristic (E) a value that is a function of values (Ei) adopted by E for each value or category (Ai) of another state characteristic (A), weighted by the overall confidence level Prob(Ai) associated with this value or category (Ai). For example, it is possible to have the formula E=Σi(Prob(Ai)*f(g(Ai), Bk))/Σi(Prob(Ai)), where f is a model or algorithm and Bk is one or more other characteristics, and Ei=f(g(Ai), Bk), g(Ai) is in particular a discrete value.

[0063] For example, the state characteristic is the metabolism level MET of the passenger, this level MET being equal to 60%, this being the confidence level, of the level MET defined for a Male and 40% of the level MET defined for a Female.

[0064] In one example, the state characteristic RClo is the level of clothing coverage of the arms of the passenger, the level RClo being equal to 50% of a level RClo associated with a long-sleeved garment and 50% of a level RClo associated with a short-sleeved garment and 0% of a level associated with bare arms.

[0065] In one example, the state characteristic is the wellbeing level of the passenger, this wellbeing level being equal to 50%, which is the confidence level, of an “Angry” level, 0% of a “Happy” level, 30% of a “Scared” level and 20% of an “Anxious” level. In this case, the state characteristic may adopt four possible categories.

[0066] In one exemplary embodiment of the invention, the system 1 comprises a plurality of measurement chains for determining values of the state characteristic, each measurement chain being designed to give values for the state characteristic with an associated confidence level (Prob(Aij)) such that the system assigns each value (Ai) an overall confidence level (Prob(Ai)) that is a function of the confidence levels (Conf(Cj)) specific to each measurement chain and of the confidence levels (Prob(Aij)) associated with each value (Aij) given by the measurement chain. The overall confidence level (Prob(Ai)) may thus be expressed as follows: Prob(Ai)=Σj[Conf(Cj)*Prob(Aij)]/Σj[Conf(Cj)]. If the values (Aij) of the characteristic (A) are numerical and continuous, the value (A*) assigned to the state characteristic (A) may then be expressed by A*=Σi(Prob(Ai)*Ai)/Σ(Prob(Ai)).

[0067] The thermal management takes into account the above values weighted by the confidence levels, in particular so as to activate thermal actuators, such as for example the HVAC or radiant panels in the vehicle.