Summary of the invention
In view of this, the problems of the prior art are directed to, the present invention is intended to provide a kind of air conditioning control method and its air-conditioning
Controller, it is mentioned above to solve the problems, such as.
The invention proposes a kind of air conditioning control methods, comprising the following steps: obtains the finger print information of user, opens empty
It adjusts;It identifies the instruction in the voice of the user, controls air-conditioning.
Further, the instruction in the voice of the identification user includes: the corresponding use of the identification finger print information
Family attribute information identifies the instruction in the voice according to the attribute information;If not identifying the finger in the voice
It enables, then the keyword in user speech is identified, according to the corresponding instruction of the Keywords matching, revision directive.
Further, the voice of the user is stored in advance, it is and described by the phonetic storage into sound bank
Fingerprint is corresponding.
Further, the customer attribute information includes pre-stored control parameter.
Further, the instruction in the voice of the identification user specifically includes: the temperature ginseng of voice selecting air-conditioning
Number, operating mode and/or wind speed, wind direction.
Another object of the present invention discloses a kind of air-conditioner controller, comprising: receiving unit, for obtaining the finger of user
Line information;Processing unit, the finger print information for the user according to acquisition are handled, and obtain control parameter set by user;
Control unit works for controlling air conditioner according to the instruction of the processing unit.
Further, the processing unit further include: analytical unit, the corresponding user of the finger print information belongs to for identification
Property information, identifies the instruction in the voice according to the attribute information;If not identifying the instruction in the voice,
The keyword in user speech is identified, according to the corresponding instruction of the Keywords matching, revision directive.
Further, the processing unit further includes storage unit, in advance to the fingerprint of the user, the attribute information
It is stored.
Further, the processing unit further includes recognition unit, for identification the instruction in the voice of the user.
Further, the storage unit is flash storage;The air-conditioner controller further includes power supply unit, described
Flash storage is connected with for the power supply unit for saving finger print data that has a power failure suddenly.
A kind of air conditioning control method and its air-conditioner controller provided by the invention have the advantages that
Fingerprint recognition, voice control are combined into control air-conditioning, the suspension of intelligent Service is carried out suitable for home environment
Formula air-conditioning and non-suspension type air-conditioning.The finger print information and voice messaging of user are extracted, carries out fingerprint recognition and voice control, simply
Fast, the time one by one by key assignments is eliminated.The present invention is also added into storage unit, the reference record that user speech is set
For customer attribute information, it is stored in storage unit, in booting next time, does not need gradually to set again, the parameter prestored may be selected, no
The parameter of setting air-conditioning is wasted time again.
For the above and related purposes, one or more embodiments include being particularly described below and in claim
In the feature that particularly points out.Other benefits and novel features will become obviously as following detailed description considers, institute
Disclosed embodiment is all such aspects to be included and they are equal.
Specific embodiment
In the following detailed description, a large amount of specific details are proposed, in order to provide a thorough understanding of the present invention.But
It will be understood by those within the art that implementable present invention without these specific details.
Below with reference to the accompanying drawings and preferred embodiment, the present invention will be described in detail.
Fig. 1 shows a kind of flow chart of air conditioning control method.
As shown in Figure 1, disclosing a kind of air conditioning control method, comprising the following steps:
S1, the finger print information for obtaining user open air-conditioning;
S2, the identification user voice in instruction, control air-conditioning.
Wherein, fingerprint recognition, voice control are combined into control air-conditioning, intelligent Service is carried out suitable for home environment
Suspension type air-conditioning and non-suspension type air-conditioning.The finger print information and voice messaging of user are extracted, fingerprint recognition and voice control are carried out,
It is simple and fast, eliminate the time one by one by key assignments.
In some illustrative embodiments, the instruction in the voice of the identification user includes: the identification fingerprint
The corresponding customer attribute information of information identifies the instruction in the voice according to the attribute information;If not identifying institute
Instruction in predicate sound then identifies the keyword in user speech, and according to the corresponding instruction of the Keywords matching, amendment refers to
It enables.
Wherein, what the identification of the finger print information carried out first is that key point just matches, and choosing fingerprint central point is key
Characteristic point, and two nearest characteristic points are found on fingerprint image, record its direction θ, position (x, y), type (breakpoint or fork
Point) and two characteristic points cross over ridge count.Compare two images corresponding informance, so that it is determined that these three characteristic points whether
Match.If matching, records key point;And so on, 3 groups of matching characteristic points are found out to and as reference point.Secondly, root
According to the reference point provided, angle twiddle factor Δ θ and shift factor (i.e. the translational movement of transverse and longitudinal coordinate) are calculated, by two width
Fingerprint image is registrated, and to the matching characteristic point in images after registration to progress similarity calculation, and is finally determined if
Matching.
Speech recognition mainly includes the pretreatment of voice signal, feature extraction, feature modeling or establishes reference model library, phase
Like these modules such as property measurements, detailed process is as follows: voice signal is converted into electric signal through microphone and is added to identifying system
Input terminal, it first has to by pretreatment, and pretreatment includes preemphasis, adding window, framing and end-point detection.After pretreatment,
It extracts the characteristic parameter of voice signal and the time series of these characteristic parameters, these data just constitutes voice to be identified
It is stored in the reference voice in storage unit with prior typing and is compared one by one by mode, and maximum similarity is to know
Other result.
Speech signal pre-processing mainly increase to voice the processing of frequency, while carrying out framing to voice, using
Start-stop point algorithm characterized by short-time energy amplitude, i.e. classical Lawrence Rabiner end-point detecting method;Due to not
When sending out sound same with speaker, perhaps the position of formant can change, but the change curve for the envelope and energy composed
It will not substantially change, therefore choose dynamic time warping (DTW) model based on Mel frequency cepstral coefficient in speech recognition algorithm
Carry out the feature extraction and identification of voice;Specific as follows: analog voice signal becomes digital signal after completing A/D conversion, to this
Time-domain signal is converted, and the time shaft of voice signal to be identified is unevenly distorted and is bent, feature and template are made
Feature alignment is calculated by constantly carrying out the smallest coupling path of vector distance between the two, and extracts MFCC parameter, to obtain
The smallest class of cumulative distance is recognition result when obtaining two vector matchings.Because of region difference, the reasons such as dialect influence, selection
HMM model is devised for the stronger speech recognition system of continuity, it is retouched jointly by two random processes that are mutually related
The statistical property of signal is stated, one of them is hidden (unobservable) Markov chain with finite state, the other is
The random process (observable) of observation vector associated with each state of Markov chain.The time varying signals such as voice are a certain
The feature of section is just described by the random process of corresponding states observation symbol, and signal changes with time by hidden Markov chain
Transition probability description.
In some illustrative embodiments, the voice of the user is stored in advance, by the phonetic storage to language
It is corresponding with the fingerprint in sound library.
Wherein, two kinds of voices of a fingerprint typing, the answer of " affirmative " and " negative ", after fingerprint typing success, system meeting
Prompt carries out the voice input that " affirmative " is answered, at this time can the continuous common particular person voice of three kinds of typing, progress is " no after the completion
The voice input answered, mode are same as above calmly ".After typing success, tone testing will be carried out, typing person records before needing to roar
Any voice (" affirmative " and " negative " will be tested) entered carries out sound bank and is successfully established confirmation if being successfully tested.
At this point, system can automatically store it into fingerprint base and sound bank respectively, to carry out when fingerprint input and when voice answering
Identification compares.
In some illustrative embodiments, the customer attribute information includes pre-stored control parameter.
Wherein, the pre-stored control parameter include user before this opens air-conditioning to airconditioning control parameter
Setting and voice messaging, finger print information.
In some illustrative embodiments, the instruction in the voice of the identification user specifically includes:
Temperature parameter, operating mode and/or wind speed, the wind direction of voice selecting air-conditioning.
Wherein, the temperature parameter of the air-conditioning includes selection refrigeration or heating and leaving air temp;Operating mode includes economize on electricity
Mode, Night, lunch break mode etc..Instruction in the voice of user further includes being timed to the air-conditioning work time, different
Period, the setting, such as energy-saving mode, Night, lunch break timing mode etc. of meeting polling mode, according to time change
It can modify to the operating mode of air-conditioning, it, will be certainly when voice prompting is continuously unresponsive if user leaves or exception occurs in air-conditioning
It is dynamic to close air-conditioning, report closing information is sent to the mobile device of user in time by communication module.In addition to this, polling
Setting of the user for mode can also grasp the situation of old man or child in family indirectly, such as old man, the child in family do not have
There is timely response, air-conditioning is closed, and is sent to user by communication module, is grasped information in family, ensures household safe.With technology
Application development in smart home, it is contemplated that it is uniformly applied to the families such as air-conditioning, TV, gate inhibition, residence abnormal monitoring
In electric appliance, home intelligence can further improve.
In some illustrative embodiments, as shown in Fig. 2, disclosing a kind of air-conditioner controller 200, comprising:
Receiving unit 201, for obtaining the finger print information of user;
Processing unit 202, the finger print information for the user according to acquisition are handled, and obtain control ginseng set by user
Number;
Control unit 203 works for controlling air conditioner according to the instruction of the processing unit 201.
Wherein, the receiving unit includes remote controler or indoor unit control panel.
In some illustrative embodiments, as shown in figure 3, the processing unit 202 further include: analytical unit 205 is used for
It identifies the corresponding customer attribute information of the finger print information, the instruction in the voice is identified according to the attribute information;If
It does not identify the instruction in the voice, then identifies the keyword in user speech, it is corresponding according to the Keywords matching
Instruction, revision directive.
In some illustrative embodiments, as shown in figure 3, the processing unit 202 further includes storage unit 206, in advance
Fingerprint, the attribute information of the user are stored.
In some illustrative embodiments, it as shown in figure 3, the processing unit 202 further includes recognition unit 204, is used for
Identify the instruction in the voice of the user.
In some illustrative embodiments, the storage unit 206 is flash storage;As shown in Fig. 2, the air-conditioning
Controller 200 further includes power supply unit 207, and the flash storage is connected with to be saved described in finger print data for having a power failure suddenly
Power supply unit 207.Wherein, the power supply unit 207 and the receiving unit 201 in air-conditioner controller 200, processing unit 202, control
Unit 203 processed connects, and powers for each unit.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.