DE102005006312B4 - Method for controlling a ventilation and/or air conditioning unit - Google Patents

Abstract

Method for controlling a ventilation and/or air conditioning unit (3) for treating interior air in a closed space (2) of a vehicle (1), in particular a motor vehicle (1), with at least two sensors (S1, S2, S3) for Detection of different chemical components of an air flow to be examined, of which a first sensor (S1) responds with higher selectivity to carbon monoxide and/or hydrogen and a second sensor (S2) with higher selectivity to hydrocarbons, and with a processor unit (P) which based on signals from the sensors (S1, S2, S3), a control signal for activating the ventilation and/or air conditioning unit (3) is generated, with the air flow to be examined being tapped from the interior air of the closed room (2), with the processor unit (P ) contains an algorithm that calculates a characteristic for a specific state from the signals of all sensors (S1, S2, S3), since s is characteristic of the current composition of the examined indoor air and is compared with the processor unit (P) stored, predetermined characteristics for a specific state for different scenarios of a possible air composition, and the algorithm has a qualitative part, with which from the signals of Sensors (S1, S2, S3) the characteristic for a specific state is calculated and compared with the specified characteristics for a specific state, and that the algorithm also includes a quantitative part with which the intensities and/or variances of the signals are determined, from the calculated characteristic for a specific state, together with the intensities and/or variances, using a look-up table, an assignment is made to a stored, specified air quality value and from this the control signal for controlling the ventilation and/or air conditioning g unit (3) is derived, characterized in that at least three sensors (S1, S2, S3) are provided, of which the third sensor (S3) responds with higher selectivity to nitrogen oxides, that with the qualitative part of the algorithm the calculated characteristic for a specific state is assigned a weighting factor , which is applied to the intensities and/or variances determined with the qualitative part of the algorithm, whereby a weighted parameter is obtained, by means of which the assignment to the air quality value from the look-up table is made, that the weighting factor from a statistically determined hedonic perception of a group of Test subjects is obtained and one of the stored, predetermined characteristics for a specific state is assigned, that a specific, predefined range of the overall parameter determined is assigned a specific, predefined control signal from a finite number n of control signals, and d in the quantitative part of the algorithm, the intensities are multiplied by the variances of the measured signals and a quantitative sum parameter is formed from this.

Description

The invention relates to a method for controlling a ventilation and/or air conditioning unit for the treatment of interior air in a closed space of a vehicle, in particular a motor vehicle, with at least 2 sensors for detecting different chemical components of an air flow to be examined, of which a first sensor responds with higher selectivity to carbon monoxide and/or hydrogen and a second sensor with higher selectivity to hydrocarbons, and with a processor unit which, based on signals from the sensors, generates a control signal for controlling the ventilation unit, with the air flow to be examined from the interior air of the closed room is tapped, the processor unit containing an algorithm that calculates a characteristic for a specific state from the signals of all sensors, which is characteristic of the current composition of the examined indoor air and is compared with predetermined characteristics stored in the processor unit for a specific state for different scenarios of a possible air composition, and wherein the algorithm has a qualitative part with which the characteristic for a specific state is calculated from the signals of the sensors and with the predetermined ones Characteristics for a specific state is compared, and that the algorithm also includes a quantitative part with which the intensities and / or variances of the signals are determined, from the calculated characteristic for a specific state together with the intensities and / or variances over a Lookup table made an assignment to a stored, predetermined air quality value and from this the control signal for controlling the ventilation and / or air conditioning unit is derived.

Such a system is described in EP 1 422 089 A2 .

Similar devices, known for example from EP 0 757 632 B1 , have long been used for pollutant-controlled ventilation of vehicle cabins in motor vehicles with switchover to recirculation mode. For this purpose, sensors measure the current pollutant content or smell of the outside air and, depending on the sensor signals, either switch to recirculation mode if the outside air is too dirty, or fresh air is supplied if the outside air is "good".

From the DE 43 28 218 A1 is known to use a sensor that responds to nitrogen oxides with higher sensitivity.

In addition, in the EP 1 256 470 B1 proposed making the response sensitivity of the changeover device for air recirculation mode dependent not only on the pollutant values measured in the outside air but also on the number of vehicle occupants, the interior temperature, the humidity in the interior, the speed of the vehicle, the fan speed of the air conditioning system, its air filter function or the compressor function. This is intended to prevent the windows from fogging up, even when the outside air quality is taken into account. However, only the current quality of the outdoor air is taken into account here, while the quality of the indoor air is completely ignored.

The device according to the above-mentioned EP 1 422 089 A2 on the other hand, enables improved ventilation and climate management of the interior atmosphere of the vehicle. By processing sensor signals for the indoor air instead of the outdoor air, it is possible to "focus" specifically on the air quality data relating purely to the indoor space. The enormously increased local reference of the measured values to the actual air quality in the passenger compartment of the vehicle opens up completely new possibilities for a significantly improved control of the ventilation and air conditioning units in closed vehicles that is precisely tailored to the current circumstances. As a result, it is possible to react to a series of standard situations (of any size and which can be specified) with very specific measures to improve the air quality in the vehicle interior. In contrast, the object of the present invention is to use the simplest possible technical means to further develop a control module of the type described at the outset such that the greatest possible reduction in the amount of data to be processed is made possible.

According to the invention, this object is achieved in a way that is as surprisingly simple as it is effective in that at least three sensors are provided, of which the third sensor responds with higher selectivity to nitrogen oxides, that with the qualitative part of the algorithm the calculated characteristic for a specific state weighting factor is assigned, which is applied to the intensities and/or variances determined with the qualitative part of the algorithm, whereby a weighted parameter is obtained, by means of which the assignment to the air quality value from the lookup table is produced that the weighting factor is obtained from a statistically determined hedonic perception of a group of test subjects and is assigned to one of the stored, specified characteristics for a specific state, that a specific, specified range of the determined overall parameter is assigned a specific, specified control signal from a is assigned to a finite number n of control signals, and that in the quantitative part of the algorithm the intensities are multiplied by the variances of the measured signals and a quantitative sum parameter is formed from this.

The "sensors" mentioned are not necessarily to be understood as a large number of physically existing individual sensors, but can also be implemented in the sense of "sensor functions", for example through different operating modes of a single physical sensor element, which provides several independent pieces of information that are related to the measurement data of several different sensor elements are comparable (keyword: "virtual sensors"). On the other hand, several differently reacting sensor elements can also be arranged spatially adjacent to one another on a common circuit board in compact arrangements.

The "air flow" does not necessarily require a physical flow of air, but the detection of the chemical composition of the indoor air can also take place in a pure diffusion mode.

Finally, a "characteristic for a specific state" can require complex pattern recognition in the case of sophisticated and technically sophisticated implementations of the invention, but in the case of simpler variants, for example, a comparison of the slope of curves of the measured signals is sufficient.

The quality of the interior air to be influenced with the control module according to the invention is adapted to the special needs of people in general or also specially selected groups of people in that the weighting factor is obtained from a statistically determined hedonic perception of a group of human test subjects and one of the stored, predetermined characteristics for one assigned to a specific state.

A further improvement is achieved in that the intensities are multiplied by the variances of the measured signals in the quantitative part of the algorithm and a quantitative sum parameter is formed from this.

A particularly high data reduction is achieved by specifying a discrete number n of control stages.

An embodiment of the method according to the invention is particularly preferred, which is characterized in that the number n of predefined control signals is selected in a range of 3≦n≦20, preferably n=5.

In practice, it proves particularly advantageous if the quantitative sum parameter is formed by multiplying the greatest intensity value of all signals by the greatest variance value of all signals.

1. a) Tabakrauch und/oder
2. b) Bioeffluenten und/oder
3. c) Lebensmittelgerüche und/oder
4. d) landwirtschaftliche Gerüche und/oder
5. e) Emissionen der Innenausstattung des Fahrzeugs und/oder
6. f) Emissionen von Mikroorganismen und/oder
7. g) gasförmige Substanzen, typischerweise aus Druckgasbehältern.

In detail, the predefined characteristics stored in the processor unit for a specific state can include the following different scenarios of a possible composition of the air:

1. a) Tobacco smoke and/or
2. b) Bioeffluents and/or
3. c) food odors and/or
4. d) agricultural odors and/or
5. e) emissions from the interior of the vehicle and/or
6. f) emissions of microorganisms and/or
7. g) gaseous substances, typically from pressurized gas containers.

This captures most of the important set pieces.

In order to not only have to reduce the ventilation and air conditioning management of the interior air to the decision "fresh air supply yes/no", in particularly preferred embodiments of the invention the controlled ventilation unit can include one or more devices for actively cleaning the interior air and the control signal of the control module as a function activate at least one of these devices based on the characteristic calculated from the signals of all sensors for a specific state.

In particular, in developments of these embodiments, the control signal can activate an activated carbon filter and/or an ozonization device and/or a device for UV catalysis on semiconductor oxide surfaces and/or an electrostatic filter and/or a device for ionizing air.

Further features and advantages of the invention result from the following detailed description of exemplary embodiments of the invention with reference to the figures of the drawing, which show details essential to the invention, and from the claims. The individual features can each be implemented individually or together in any combination in variants of the invention.

The schematic drawing shows exemplary embodiments of the invention, which are explained in more detail in the following description.

• 1 eine schematische Darstellung eines Fahrzeugs mit Steuermodul zur Ansteuerung einer Lüftungs- und Klimatisierungseinheit; und
• 2 ein Verlaufsdiagramm zur Berechnung einer Luftgütestufe aus den Sensordaten.

Show it:

• 1 a schematic representation of a vehicle with a control module for controlling a ventilation and air conditioning unit; and
• 2 a progress chart for calculating an air quality level from the sensor data.

In 1 one recognizes a vehicle 1, the interior air of which is conditioned in the closed interior 2 by means of a ventilation and/or air conditioning unit, of which only a ventilation 3 is indicated schematically in the drawing. The ventilation and/or air conditioning unit is activated via control signals, which are generated by a control module modified to carry out the method according to the invention, which includes a processor unit P and sensors S1, S2, S3 for examining the interior air in the closed space 2 of the vehicle 1. The sensors S1, S2, S3 are again shown only very schematically. In the exemplary embodiment shown, sensor S1 should be particularly sensitive to carbon monoxide and/or hydrogen, sensor S2 should be particularly sensitive to hydrocarbons, and sensor S3 should be particularly sensitive to nitrogen oxides. The processor unit P in this exemplary embodiment works with an algorithm that includes an electronic filter F, for example a median filter for removing artefacts and smoothing the measurement curve and for preparing a baseline correction.

The algorithm uses the signals from the sensors S1, S2, S3 for the interior air to calculate a characteristic for a specific condition, which is characteristic of the current composition of the indoor air being examined, and with predefined characteristics stored in the processor unit P for a specific condition for different scenarios a possible air composition is compared.

The internal air quality in vehicle 1 can only be improved by measures such as air exchange and active purification. Efficient use of appropriate measures presupposes that the status of the air quality in the interior 2 of the vehicle 1 is precisely known.

• 1.Tabakrauch
• 2.Bioeffluenten
• 3.Lebensmittel
• 4.Landwirtschaft
• 5.Innenausstattungsemissionen
• 6.Schimmel

The multi-gas sensor module modified to carry out the method according to the invention is intended to provide an internal air quality factor within the passenger compartment in relation to the presence and concentration of odorous substances such as cigarette smoke, putrid/moldy smells, liquid manure and emissions caused by occupants. The significance of the scenarios is broken down as follows, taking feasibility into account:

• 1.Tobacco smoke
• 2.Bioeffluents
• 3.Food
• 4.Agriculture
• 5.Interior emissions
• 6.mold

Depending on the air quality factor, the parameters of the automatic climate control should be adjusted. The goal is to improve the internal air quality.

• 1. Hohe Luftgüte innen -hohe Luftgüte außen keine Maßnahme
• 2. Hohe Luftgüte innen -schlechte Luftgüte außen (z.B. Tunnelfahrt oder Innenstadtverkehr)
  • ⇨ Umluftbetrieb, aktive Reinigung von Außenluft und Zumischung bei lang anhaltenden Ereignissen wg. steigender CO₂-Konzentration
• 3. Schlechte Luftgüte innen -hohe Luftgüte außen (z.B. Rauchen auf einer Landstraße)
  • Luftwechselrate erhöhen
• 4. Schlechte Luftgüte innen -schlechte Luftgüte außen (z.B. Rauchen im Tunnel)
  • ⇨ Umluftbetrieb mit aktiver Reinigung, bei lang anhaltenden Ereignissen Zumischung von Außenluft

In simplified terms, the following cases can be distinguished:

• 1. High air quality inside - high air quality outside no measure
• 2. High air quality inside - poor air quality outside (e.g. tunnel driving or inner-city traffic)
  • ⇨ Air recirculation mode, active cleaning of outside air and admixture in the event of long-lasting events due to increasing CO ₂ concentration
• 3. Poor air quality inside - high air quality outside (e.g. smoking on a country road)
  • Increase air exchange rate
• 4. Bad air quality inside -poor air quality outside (e.g. smoking in the tunnel)
  • ⇨ Air recirculation mode with active cleaning, addition of outside air in the event of long-lasting events

multi-gas sensor approach

• • Externe Emissionsquellen
  • ◯ Gülle (Phenole, Ketone, Aldehyde, Fettsäuren, Amine etc., komplexe Mischung mit individuellen Konzentrationen im ppb-Bereich)
• • Interne Emissionsquellen
  • ◯ Tabakrauch
  • ◯ Bioeffluenten
  • ◯ Lebensmittel
  • ◯ Innenausstattungsemissionen
  • ◯ Schimmel

Experience with "electronic noses" collected in the past shows that a classification into good and bad smells is hardly possible. This follows from the fact that no clear correlation can be established between the chemical structure of a substance and its hedonic effect. Nonetheless, there is a potential way to detect unpleasant odor events related to:

• • External emission sources
  • ◯ Liquid manure (phenols, ketones, aldehydes, fatty acids, amines, etc., complex mixtures with individual concentrations in the ppb range)
• • Internal emission sources
  • ◯ Tobacco smoke
  • ◯ Bioeffluents
  • ◯ Food
  • ◯ Interior emissions
  • ◯ mold

For this purpose, the respective signal pattern is compared with the stored patterns that are typical for the specific scenarios and assigned on the basis of its similarity.

sensors

MOS sensors generally consist of a substrate (eg ceramic or Si) which is coated with the sensitive material, often tin dioxide. In order to achieve different selectivities, the tin dioxide is doped with precious metals (Pd, Pt, Au etc.) in different concentrations. The term "doping" has become established in the literature, but the effect of the precious metals is mainly catalytic). In addition, the substrate is heated to temperatures between 200 and 450°C. The oxygen species present on the surface of the MOS sensor react with components from the gas phase and lead to a change in the concentration of free charge carriers in the n-type metal oxide. Reducing gases (eg CO) lead to a decrease and oxidizing gases (eg NO ₂ ) to an increase in resistance. Non-linear behavior can very often be observed, which can be compensated for in the subsequent data treatment (“forced linearization”).

The different substrate technologies are listed below and the advantages and disadvantages of each method are presented: Micromechanical substrate: Materials frame: silicon Material membrane: - Electrode Material: mostly platinum heater material: mostly platinum substrate temperature: Si frame approx. 50°C (at membrane temperature of 350°C) Power Consumption: approx. 20-50mW (at operating temperature of 350°C) Packaging: COB (chip-on-board) or on socket (e.g. TO 39) Coating: drop methods Advantages: -low substrate frame temperature with optimal membrane temperature at the same time -low power consumption -flexible packaging options through the use of different materials or technologies and thus significant cost savings -long-term stable - State of the art (Si technology) Disadvantage: -relatively new technology (about 5 years)

algorithm

The evaluation algorithm is the heart of the multi-gas sensor, as it imparts a certain "intelligence" to the sensor signals. The sensor signal, which depends on the presence of gaseous substances, must be transformed into an air quality level (AQL) that is related to human perception/sensation. Due to the hedonic reference to the air quality level, the approach goes beyond the purely linear conversion of physical variables. Due to different odor perception (odor threshold) and association, the same concentrations of different chemical substances are evaluated very differently. It is therefore necessary to recognize defined scenarios based on their signal pattern in order to give the output signal the desired hedonic weighting.

• Transformation der Widerstandswerte in die Leitfähigkeit
• Sensoranpassung durch einen Kalibrierfaktor c, damit alle Module das gleiche Verhalten zeigen. $x_{transferSensori}=x_{Sensori}\times c_{Sensori}$
  
• Median-Filter zum Entfernen von Artefakten und Glättung der Messkurve Basislinienkorrektur, um die zeitliche Fluktuation der Basislinie auszugleichen.
• Standardisierung der Variablen, damit die Daten der unterschiedlichen Sensoren relativ zu ihrer Detektionsgrenze (LOD) gleich gewichtet werden.

$$x_{stdSensori}=\frac{x_{blcSensori}}{c_{Sensori}}$$

$$c_{Sensori}=5\times \sigma$$

• Intensitätsberechnung durch Projektion der dreidimensionalen Daten auf eine Gerade. Hierzu wird jeweils zu jedem Messzeitpunkt i die größte standardisierte Variable der drei Sensoren als Intensität ausgegeben.

$$Intensit\ddot{a}t\left(i\right)=Max\left(x_{stdSensor1}\left(i\right),x_{stdSensor2}\left(i\right),x_{stdSensor3}\left(i\right)\right)$$

• Varianzabschätzung für jeden Sensor

$$e\left(i\right)=x_{Sensori}\left(i\right)-m\left(i-1\right)$$

$$m\left(i\right)=m\left(i-1\right)+{\mu }_m\cdot e\left(i\right)$$

$$V\left(i\right)=V\left(i-1\right)+{\mu }_V\cdot e\left(i\right)$$

$$Vm\left(i\right)=Vm\left(i-1\right)+{\mu }_{Vm}\left(V\left(i-1\right)-Vm\left(i-1\right)\right)$$

wobei:

• e(i) das Fehlersignal
• m(i) der geschätzter Mittelwert
• V(i) die geschätzte Varianz
• Vm(i) die geschätzte mittlere Varianz

A possible, preferred procedure for calculating an air quality level from the sensor data is in 2 shown. The individual steps are as follows:

• Transformation of resistance values into conductivity
• Sensor adjustment by a calibration factor c, so that all modules show the same behavior. $x_{t\mathrm{right}ansfe\mathrm{right}SensO\mathrm{right}i}=x_{SensO\mathrm{right}i}\times c_{SensO\mathrm{right}i}$
  
• Median filter to remove artifacts and smooth the trace. Baseline correction to compensate for baseline fluctuations over time.
• Standardization of the variables so that the data from the different sensors are weighted equally relative to their limit of detection (LOD).

$$x_{st\mathrm{i.e}SensO\mathrm{right}i}=\frac{x_{blcSensO\mathrm{right}i}}{c_{SensO\mathrm{right}i}}$$

$$c_{SensO\mathrm{right}i}=5\times \sigma$$

• Intensity calculation by projecting the three-dimensional data onto a straight line. For this purpose, the largest standardized variable of the three sensors is output as intensity at each measurement time i.

$$Intensit\ddot{a}t\left(i\right)=Max\left(x_{st\mathrm{i.e}SensO\mathrm{right}1}\left(i\right),x_{st\mathrm{i.e}SensO\mathrm{right}2}\left(i\right),x_{st\mathrm{i.e}SensO\mathrm{right}3}\left(i\right)\right)$$

• Variance estimation for each sensor

$$e\left(i\right)=x_{SensO\mathrm{right}i}\left(i\right)-m\left(i-1\right)$$

$$m\left(i\right)=m\left(i-1\right)+{\mu }_m\cdot e\left(i\right)$$

$$V\left(i\right)=V\left(i-1\right)+{\mu }_V\cdot e\left(i\right)$$

$$Vm\left(i\right)=Vm\left(i-1\right)+{\mu }_{Vm}\left(V\left(i-1\right)-Vm\left(i-1\right)\right)$$

whereby:

• e(i) the error signal
• m(i) the estimated mean
• V(i) the estimated variance
• Vm(i) the estimated mean variance

Normalized Variance

$$V_{nO\mathrm{right}m}\left(i\right)=\sqrt{\frac{V\left(i\right)}{Vm\left(i\right)}}$$

Like the intensity calculation, the max. variance is determined by projecting the three-dimensional data onto a straight line. $$Max.Va\mathrm{right}ian\mathrm{e.g}\left(i\right)=Max\left(V_{nO\mathrm{right}mSensO\mathrm{right}1}\left(i\right),V_{nO\mathrm{right}mSensO\mathrm{right}2}\left(i\right),V_{nO\mathrm{right}mSensO\mathrm{right}3}\left(i\right)\right)$$

Signal improvement for pattern recognition with weakly and undoped sensors through a new composite quantity. For this purpose, the portion of the standardized conductivity that is less than zero is replaced by the inverted resistance. This measure combines the advantages of both variables in baseline correction.

$$x_{normSensor1}=\frac{x_{stdSensor1}}{V_L};x_{normSensor2}=\frac{x_{stdSensor2}}{V_L};x_{normSensor3}=\frac{x_{stdSensor3}}{V_L}$$

$$sim=100-\sqrt{{\left(x_{normSensor1}-x_{calSensor1}\right)}^2+{\left(x_{normSensor2}-x_{calSensor2}\right)}^2+{\left(x_{normSensor3}-x_{calSensor3}\right)}^2}\times 100$$

errechnet wird.Pattern recognition is based on the principle that the data is normalized via the vector length (VL) and then a similarity (sim) in percent from the difference to predefined vectors that correspond to the scenarios $$V_L=\sqrt{x_{st\mathrm{i.e}SensO\mathrm{right}1}{}^2+x_{st\mathrm{i.e}SensO\mathrm{right}2}{}^2+x_{st\mathrm{i.e}SensO\mathrm{right}3}{}^2}$$

$$x_{nO\mathrm{right}mSensO\mathrm{right}1}=\frac{x_{st\mathrm{i.e}SensO\mathrm{right}1}}{V_L};x_{nO\mathrm{right}mSensO\mathrm{right}2}=\frac{x_{st\mathrm{i.e}SensO\mathrm{right}2}}{V_L};x_{nO\mathrm{right}mSensO\mathrm{right}3}=\frac{x_{st\mathrm{i.e}SensO\mathrm{right}3}}{V_L}$$

$$sim=100-\sqrt{{\left(x_{nO\mathrm{right}mSensO\mathrm{right}1}-x_{calSensO\mathrm{right}1}\right)}^2+{\left(x_{nO\mathrm{right}mSensO\mathrm{right}2}-x_{calSensO\mathrm{right}2}\right)}^2+{\left(x_{nO\mathrm{right}mSensO\mathrm{right}3}-x_{calSensO\mathrm{right}3}\right)}^2}\times 100$$

is calculated.

In order to limit cases of misclassification, two additional criteria were introduced. For the scenario to be specified, the similarity must be over 85% in the case of “fast food” and “slurry” and over 65% in the other cases.

Furthermore, a range for the vector length is defined for each scenario in order to filter out small or unusually large events.

Weighting factor and gain are defined depending on the detected scenario. The weighting factor (w(i)) indicates the factor by which the output signal is to be amplified. With this measure the correlation of the output signal with the hedonic effect of a scenario can be improved. The decay behavior of the AQL signal is adjusted via the gain according to the detected scenario.

New intensity of the output signal is the product of intensity, variance and weighting factor. $$Inew\left(i\right)=V\left(i\right)\cdot I\left(i\right)\cdot w\left(i\right)$$

By using two different components to generate the output signal, the dynamics of the signal are increased, drift effects are reduced and, as a result, the number of "incorrectly" detected events is reduced.

Lookup table for the air quality classes assigned to the intensities. Table 1: Example of a lookup table intensity new air quality class air pollution < 800 0 None - very weak 800-2500 1 Weak 2500-4000 2 Clearly 4000-6000 3 Strong 6000-9000 4 Very strong >9000 5 Extremely strong

Calculate signal decay of the AQL signal.

Due to the physical conditions, an exponential decrease in the concentration of gaseous substances in the interior can be assumed. Therefore, the decay of the AQL signal was adapted to this law. However, with a view to better implementation on the microcontroller, only an approximation of the exponential function was used. $$AQ{L}_{AbklinGen}\left(i\right)=\left(1-Gain\right)\cdot AQ{L}_{AbklinGenl}\left(i-1\right)$$

The Gain parameter can be used to set the signal's acceptance rate. The newly calculated AQL is of course rounded to discrete values before output. Should the current AQL value be greater than AQL. Decay is set as the new AQL _Decay .

Technical possibilities for active air purification

• ⇨ Aktivkohlefilter
• ⇨ Ozonisierung
• ⇨ UV-Katalyse an Halbleiteroxidoberflächen
• ⇨ Elektrostatische Filter
• ⇨ Ionisierung

Possible methods for air purification are:

• ⇨ Activated carbon filter
• ⇨ Ozonization
• ⇨ UV catalysis on semiconductor oxide surfaces
• ⇨ Electrostatic filters
• ⇨ Ionization

Use of active filtering

Since the cleaning performance of all currently known active filter methods is limited due to the technically integrable construction volume, only a partial air flow can be directed over the filter. Therefore, as in cases 2 and 4, the inflow of poor quality outside air must be reduced by the air recirculation mode. In both cases, however, a flow must be maintained in order to reduce the accumulation of carbon dioxide in the vehicle.

The increased local reference of the measurement data when evaluating the indoor air, which is achieved with the multi-gas sensor module according to the invention, opens up completely new possibilities for improved ventilation and climate management.

Claims (6)

Method for controlling a ventilation and/or air conditioning unit (3) for treating interior air in a closed space (2) of a vehicle (1), in particular a motor vehicle (1), with at least two sensors (S1, S2, S3) for Detection of different chemical components of an air flow to be examined, of which a first sensor (S1) responds with higher selectivity to carbon monoxide and/or hydrogen and a second sensor (S2) with higher selectivity to hydrocarbons, and with a processor unit (P) which based on signals from the sensors (S1, S2, S3), a control signal for activating the ventilation and/or air conditioning unit (3) is generated, with the air flow to be examined being tapped from the interior air of the closed room (2), with the processor unit (P ) contains an algorithm that calculates a characteristic for a specific state from the signals of all sensors (S1, S2, S3), since s is characteristic of the current composition of the examined indoor air and is compared with the processor unit (P) stored, predetermined characteristics for a specific state for different scenarios of a possible air composition, and the algorithm has a qualitative part, with which from the signals of Sensors (S1, S2, S3) the characteristic for a specific state is calculated and compared with the specified characteristics for a specific state, and that the algorithm also includes a quantitative part with which the intensities and/or variances of the signals are determined, from the calculated characteristic for a specific state, together with the intensities and/or variances, using a look-up table, an assignment is made to a stored, specified air quality value and from this the control signal for controlling the ventilation and/or air conditioning g unit (3) is derived, characterized in that at least three sensors (S1, S2, S3) are provided, of which the third sensor (S3) responds with higher selectivity to nitrogen oxides that with the qualitative part of the algorithm the calculated characteristic for a weighting factor is assigned to a certain state, which is applied to the intensities and/or variances determined with the qualitative part of the algorithm, whereby a weighted parameter is obtained, by means of which the assignment to the air quality value from the look-up table is made, that the weighting factor consists of a statistically determined hedonic perception of a group of test persons is obtained and one of the stored, predetermined characteristics for a specific state is assigned that a specific, predetermined range of the determined overall parameter has a specific, predetermined control signal from a finite number n of taxes rsignalen is assigned, and that in the quantitative part of the algorithm, the intensities are multiplied by the variances of the measured signals and a quantitative sum parameter is formed from this. procedure after claim 1 , characterized in that the following applies to the number n of predetermined control signals: 3 ≤ n ≤ 20, preferably n = 5. procedure after claim 1 or 2 , characterized in that the quantitative sum parameter is formed by multiplying the greatest intensity value of all signals by the greatest variance value of all signals. Method according to one of the preceding claims, characterized in that the predetermined characteristics stored in the processor unit (P) for a specific state include the following different scenarios of a possible air composition: a) tobacco smoke and/or b) bioeffluents and/or c) food odors and /or d) agricultural odors and/or e) emissions from the interior of the vehicle (1) and/or f) emissions from microorganisms and/or g) gaseous substances, typically from compressed gas containers. Method according to one of the preceding claims, characterized in that the controlled ventilation and / or air conditioning unit (3) comprises one or more devices for actively cleaning the indoor air, and that the control signal of the control module depending on the characteristic calculated from the signals of all sensors (S1, S2, S3) for a specific state, activates at least one of these devices if necessary. procedure after claim 5 , characterized in that the control signal can activate an activated carbon filter and/or an ozonization device and/or a device for UV catalysis on semiconductor oxide surfaces and/or an electrostatic filter and/or a device for ionizing air.

Images