WO2011018562A1

WO2011018562A1 - Electronic device for controlling electrical apparatuses by means of sound

Info

Publication number: WO2011018562A1
Application number: PCT/FR2010/000566
Authority: WO
Inventors: Richard Thibert
Original assignee: Richard Thibert
Priority date: 2009-08-10
Filing date: 2010-08-06
Publication date: 2011-02-17
Also published as: FR2949012B1; FR2949012A1

Abstract

The invention relates to an electronic device (1) for controlling one or more electrical apparatuses (13, 15) by detecting sound signals (5). The audio signal of the microphone (2) is amplified and digitized so as to be compared with the signatures (4) of the expected control signals (3) such that exceeding a threshold leads to detecting and controlling (12, 14) for turning on, turning off or changing the controlled electrical apparatus(es) (13, 15). The sound control signals (6) in the form of audio files (9) may thus be used by any multimedia apparatus (5) to remotely control the apparatuses (13, 15).

Description

ELECTRONIC DEVICE FOR CONTROLLING ELECTRICAL EQUIPMENT BY SOUND

The present invention relates to the field of electrical equipment and home automation. The present invention relates more particularly to a device making it possible to control

remote electrical equipment by broadcasting sound signals. Currently to implement a home automation system there are two approaches:

- implement the installation during the construction of the habitat, the installation then uses cables and equipment "fixed" non-evolutive and installed by a professional. Very few private homes have a fixed home automation system, even in new homes or renovated it is very rare to provide a home automation system. There are many reasons for this: extra cost, obsolescence of equipment, reliability ...;

- implement on an existing installation home automation equipment (often removable).

Our invention is interested in this second situation by proposing a device that can be installed like any electrical equipment.

Such a device can not involve wiring or

changes to the existing electrical installation, so it must integrate as easily as possible in a traditional habitat using a wireless communication protocol.

Several standards, projects and systems have been devised for this purpose, but the technologies selected have their limits:

- The carrier current is not compatible with the three-phase. It has a range which depends on the electrical installation and is very disturbed by the CPL boxes, switching power supplies or low consumption lamps. So he has reality a relatively small scope (immediate "neighbor" parts) or simply does not work if the conflict with other "power line" equipment is too strong;

- the radio frequency has a reduced range taking into account

Authorized powers and is prone to parasites especially with Wifi and low energy lamps, urban environment, a simple wall on the direct path makes communication difficult on advanced protocols (WIFI, DECT). The simpler home automation protocols make it possible to offer cheaper products, but are even more affected;

- the infrared only works in direct proximity.

The most successful systems combine these different

technologies in order to limit the inconvenience of each approach without, however, completely solving the problems. In addition, these devices generally require a specific interface whose cost is added to that of the boxes. Moreover, this interface is generally only available for computers (USB connection) and not for other equipment such as mobile phones, music players ...

The accompanying drawings illustrate the invention:

- Figure 1 shows the device and its environment;

FIG. 2 illustrates possible uses of the device; Figure 3 shows the principle of convolution;

FIG. 4 represents the probability distribution of the signal after convolution in a totally noisy environment;

- Figure 5 shows the possible states and recognized command signals.

With reference to these figures, the invention is an electronic device (1) for controlling one or more electrical equipment (13, 15) from the detection of sound signals, comprising a microcontroller electronic card or programmable logic array or equivalent active components, a micro audio (2), a connection to the electrical network (10), a regulated power supply (16), a connection to the equipment controlled through a control component of the relay type, triacs, thyristors or

equivalent (12,14).

the audio signal of the microphone (2) is amplified and digitized, the signatures (4) of the control sound signals (6) are stored in non-volatile memory, at least 2 control sound signals (6) are defined for each control (12). , 14), one for its ignition and the other for its extinction, the control sound signals (6) and the signatures (4) are coupled (11) so that a signature characterizes the control signal, the detecting (3) the control signals (6) being performed by comparing the digitized audio signal (8) with the expected signatures (4) so that exceeding a threshold (17) leads to detection and control ( 12,14) for switching on, off or changing the controlled electrical equipment (13,15).

The control sound signals (6) can be set to

provision in the form of computer files (9) recorded on removable media in MP3, WAV, AIFF or equivalent formats. They can also be downloadable over the internet from a single reference indicated with the receiver (1).

The signatures can be predefined during manufacture or:

- recorded in the device with buttons for triggering the recording of the control sound signals by the microphone (2) and deduce the signature (4);

- defined by a group of switches allowing

select the family of sound signatures (4) command that will react this case.

A visual indicator can indicate the state of the device namely the position (on or off) of the controlled equipment (12,14) as well as the proper operation of the receiver (1).

A loudspeaker may be added to the receiver (1) to emit sound signals for other receiving devices or multimedia equipment. An automatic microphone gain control system (2) can be provided to adapt to the ambient sound level.

All the elements of the receiver (1) can be integrated in a housing having a male electrical outlet for connection to the electrical network and one or more female electrical outlets for connecting one or more controlled equipment.

The device can use for detection, convolution in its direct or optimized form FFT. The principle of the invention therefore consists in using the sound (sound wave) in order to transmit the commands to the devices:

- the components for receiving (2), transmitting (7) and processing

(8,3,4,17) the signal are cheap and common;

- no regulation limits the use of the media

unlike the radio waves or the carrier current (except a maximum volume not to damage the hearing, well above our needs);

- no health impact is to be feared (unlike electromagnetic radiation);

- current computer and telephone equipment,

multimedia, ... (20,21,22,23,24) allow to receive or to emit an audio signal, the interface with these equipments does not require any device or specific connection.

The receiver (1) allows the control of electrical equipment (25,26) when broadcasting specific sound signals (6).

Any device capable of generating sound (20,21,22,23,24) may control the receivers (1). Thus without any other specific equipment than the receiver (1), it is possible to trigger the switching on or off of electrical equipment:

- by reading a control MP3 from his computer (20), his player (22) or his phone (21,23) placed in the room or a neighboring room, this piece may for example turn off the light at the end of a reading list ;

- remotely by calling his cell (21) in the room with an MP3 ringer command or by calling his answering machine and broadcasting the command audio signal from his

music player ;

- with the remote control of its hi-fi system (24) by triggering the playback of a control song (recorded in MP3 or audio CD ...);

at a defined time using an alarm clock (24) which broadcasts a control piece to turn on the light for example;

- with specific software installed on a computer (20) or its mobile phone (21).

The receiver can also:

- allow the variation of a luminaire (26) by offering a command on / off or level (for example from 1 to 10);

- act as a "multiple" outlet controlled to order separately several equipment (13,15);

- integrate into a module intended for the electrical installation as a panel component or a switch.

The communication is one-way Transmitter (5) → Receiver (1):

- the transmitter (5) diffuses by its loudspeaker (7) the control audio signal (6) associated with the order to be sent;

the receiver (1) continuously listens for the audio signals by means of its microphone (2) and when it detects (3) a control signal corresponding to the signature (4) it controls (12, 14)

the power supply of the equipment (13,15).

With this solution any equipment capable of generating a sound signal: MP3 player, telephone, television, stereo ...

(20,21,22,23,24) can become transmitter (5).

The orders recognized by the receiver are (Figure 5):

- ignition of the equipment;

- extinction of the equipment;

- possibly a power level for the equipment (for example in steps of 20%).

For each order a command signal must be used. The detection (3) being reliable (see order detection below), the transmitter (5) may consider that the receiver (1) has received and understood the issued order.

An "acknowledgment" mechanism is possible by coupling a transmitter to the receiver to confirm the order received by the playback of a specific audio file.

The control signal (11) corresponds to the specific sound signal of the command considered. This control signal consists of the audio signal (6) and the signature which makes it possible to detect it (4).

The control signals must be shared between the transmitter (5) and the receiver (1):

- the transmitter (5) must be able to broadcast the audible signal, it must therefore have the sound signal in a form

"Readable" (6). For this, the sound signal is stored in a file (9) and is communicated to the transmitter (5), for example by a memory card for a player (22), by burning an audio CD for a stereo (24) transferring the file over a local area network for a computer (20), ...

the receiver (1) must have a "signature" (4) which makes it possible to recognize the corresponding sound signal.

To share the signature pair (4) - audio file (6) between the transmitter and the receiver, three possibilities exist, the same receiver can combine these three solutions or implement only one to reduce its cost:

- The receiver (1) is preconfigured during manufacture, each receiver (1) has a set of unique sound signals (a signal for each command) whose signatures (4) are already recorded. The corresponding signals are made available to the receiver either directly by a CD, a USB key ... provided with the receiver or by downloading on the internet with the reference of the receiver box.

- The receiver (1) is preconfigured with a large set of control signals, the user chooses by a group of switches the signals he wants to use. This solution is identical to the first option in principle, but allows to have all the identical receivers during their manufacture and to control several receivers (1) with the same control signal (6).

- The receiver (1) can learn to recognize a control signal, for this the user must broadcast the signal (6) and press the button corresponding to the registration of the command. In this case the signature (4) corresponds to

the recording made by the receiver (1) of the control audio signal (6).

For the preconfigured receivers (1), the control signal (11) is generated and not selected by the user. For this we can use any audio signal. We used a pseudo random signal. This type of uncorrelated signal provides a maximum detection capability. In this case, only the parameters of the signal generation must be transferred to the receiving equipment to have the signature (4), which reduces the memory space for storing the signal.

The audio signal is picked up by the microphone (2), then digitized (8) before feeding the detector (3). Since the dynamic range of the sound environment in which we live is greater than the resolution of the converter (8) and we

If you want to detect a signal that is potentially heavily embedded in noise, it is interesting to dynamically adjust the gain of the amplification of the microphone to optimize the resolution of the scan. This amplification can be global or performed by frequency band (multiband compression).

In order to recognize the sound "order", we use a "convolution" detection.

This technique consists in finding the correlation between a control signal and the received signal. The sensitivity and reliability of the detection can be increased without limit by lengthening the duration of the control signal. Thus unlike other types of detection convolution is very effective on a long command signal, in our use a fraction of a second to several seconds. This allows a quality detection even if the signal is strongly altered. Figure 3 shows the formulas of convolution C (t)

transposed into a finite domain.

In practice, the system permanently convolves (3) between the expected signatures (4) and the digitized audio signal (8) received by the microphone (2). As soon as the function of

convolution exceeds a defined threshold (17), the order associated with the signature is detected. The threshold (17) is defined statistically according to the probability of false positive that we accept.

Direct convolution is a greedy operation in computing time, to process n samples the complexity is expressed in n ² .

We have implemented several optimizations to use microcontrollers or programmable logic networks of low cost:

- The convolution is performed in the frequency space using FFT (fast Fourier transform), in this case the complexity is expressed in n.ln (n). For a reference signal of 20,000 samples the calculation time gain of the FFT on the direct calculation can be a factor of 10,000;

the signature (4) is directly expressed in the frequency domain which also optimizes the detection work;

- Depending on the state of the receiver (1) it is possible to limit the number of control signals to be detected. Figure 5 shows the expected signals according to the state of the receiver. Thus, for an ON / OFF command only a control signal must be detected: On if off and Off if on. For drive operation two control signals must be detected (C signal + 20% and D -20% signal).

We have confirmed the detection capacity of the control signal on 10000 measurements (white noise signal and signature,

16-bit quantization, computation carried out on 22,050 samples is a reference of 1 second with a sampling with 22 kHz), figure 4 corresponds to the distribution of the signal at the exit of the

detector.

From these elements, we determine for example a threshold (17) of 4.68% (expressed in% of the autocorrelation) for detection reliability admitting a false positive per year for a totally noisy environment (100% white noise). This provides a detection capacity of 13 dB or a range of 32 m for a signal of the same intensity as the ambient noise.

This result is already very good, can be increased tenfold by using a longer signal for example 10s sampled at 44.1 kHz.

We show that the invention responds well in terms of range and transmission reliability.

The device therefore provides a reliable control of any electrical equipment by broadcasting a sound signal from any multimedia equipment.

Claims

CLAIMS Electronic device (1) for controlling one or more electrical equipment (13, 15) from the detection of sound signals, comprising a card

microcontroller electronics or programmable logic network or equivalent active components, a microphone audio (2), a connection to the electrical network (10), a regulated power supply (16), a connection to equipment controlled through a relay-type control component , triacs, thyristors or equivalent (12,14) characterized in that:

the audio signal of the microphone (2) is amplified and digitized,

the signatures (4) of the control sound signals (6) are stored in non-volatile memory,

at least two control sound signals (6) are defined for each control (12, 14), one for its ignition and the other for its extinction,

the control sound signals (6) and the signatures (4) are coupled (11) so that a signature characterizes the control signal,

the detection (3) of the control signals (6) being

performed by comparing the digitized audio signal (8) with the signatures (4) expected so that the exceeding of a threshold (17) leads to the detection and control (12,14) allowing the ignition, extinguishing or the

variation of the electrical equipment ordered (13,15).

Electronic device according to claim 1, characterized in that the control sound signals (6) are made available as computer files (9)

recorded on removable media in MP3, WAV, AIFF or equivalent formats.

Electronic device according to claim 1, characterized in that the control sound signals (6) are

downloadable via the internet from a single reference. Electronic device according to claim 1, characterized in that buttons are used to trigger

recording of the control sound signals by the microphone (2) and deduce the signature (4).

5 Electronic device according to claim 1, characterized in that it comprises a group of switches for selecting the family of signatures (4).

6 Electronic device according to claim 1, characterized in that it comprises, a visual indicator indicating its state. Electronic device according to claim 1, characterized in that it comprises a loudspeaker for emitting sound signals for other devices or equipment

multimedia.

8 electronic device according to claim 1, characterized in that it comprises, an automatic microphone gain control system (2) to adapt to the ambient sound level.

9 Electronic device according to claim 1, characterized in that it is integrated in a housing having a male electrical outlet for connection to the electrical network and one or more female electrical outlets for connecting one or more controlled equipment.

Electronic device according to claim 1, characterized in that the convolution in its direct or optimized FFT form is used for detection.