CN109927072A - Speech robot department of human head and neck device and control for the inquiry of factories and miness hospital department - Google Patents

Abstract

Belong to intellect service robot field for the speech robot department of human head and neck device of factories and miness hospital department inquiry and control；It solves the problems, such as the interactive difference of guest-meeting robot in the prior art and does not have department's query function；Including robot head, human-computer dialogue device, neck device and control system；In neck device upper end, the human-computer dialogue device is arranged inside robot head for the robot head setting；The robot head includes holder,head and blink mechanism；The human-computer dialogue device includes phonetic incepting plate, action control plate, loudspeaker, horn bracket and Mike；The neck device includes neck swing mechanism and neck luffing mechanism；For the present invention while realizing conventional welcome's function, human-computer dialogue device can answer the specific department position of clients inquiry, and can complete to blink when voice answering is inquired, nod, these interactivity movements of shaking the head, and be applied to factories and miness hospital welcome field.

Description

Speech robot department of human head and neck device and control for the inquiry of factories and miness hospital department

Technical field

The invention belongs to intellect service robot field more particularly to a kind of phonetic machines for the inquiry of factories and miness hospital department Device department of human head and neck device and control.

Background technique

As robot technology is gradually improved, robot gradually starts to come into service industry, as service class robot point The guest-meeting robot of branch, and single duplicate mechanical action of conforming to the principle of simplicity is promoted to development has height intelligence, initiative and can To realize the anthropomorphic robot of interactive voice, voice be not only it is interpersonal carry out information exchange most directly, most convenient and The important tool exchanged between most effective tool and people and robot, in past twenty or thirty year, graphical user circle Face (GUI) has been achieved for very big progress and development, but is not still got rid of using GUI and operated with hand, therefore, voice For user interface (VUI) at next target of people, it is desirable to can issue to instruct to robot by sound.

The present invention is applied to factories and miness hospital welcome field, and since factories and miness hospital services face is big, especially retired personnel is more, It is overstaffed between date, for lacking the factories and miness hospital services hall of professional hospital guide personnel, personnel are noisy will increase it is medical The psychological pressure of person brings bad medical experience and is delayed effective interrogation time, in the prior art the neck of guest-meeting robot Part device cannot complete blink, point when can not achieve the specific department position for answering clients inquiry and voice answering inquiry The problem of head, these interactivities of shaking the head act.

Summary of the invention

The present invention overcomes the deficiency in the above-mentioned prior art, the speech robot people for the inquiry of factories and miness hospital department is provided Neck part device and control, while realizing conventional welcome's function, human-computer dialogue device can answer the tool of clients inquiry Body department position, and can complete to blink when voice answering is inquired, nod, these interactivity movements of shaking the head, solve existing skill Guest-meeting robot is interactive poor in art and does not have the problem of department's query function, and alleviates factories and miness hospital diagnosis-guiding personnel's Operating pressure improves the medical experience of clients, improves the overall image of hospital.

Technical solution of the present invention:

For the speech robot department of human head and neck device of factories and miness hospital department inquiry, including robot head, human-computer dialogue device, neck Part device and control system；In neck device upper end, the human-computer dialogue device is embedded in robot for the robot head setting Head；The robot head includes holder,head and blink mechanism；

The human-computer dialogue device includes phonetic incepting plate, action control plate, loudspeaker, horn bracket and Mike；The neck Device includes neck swing mechanism and neck luffing mechanism；The neck elevation mount is arranged in robot head lower end；

The set-up of control system is in human-computer dialogue device；The control system includes the first chip and the second chip；It is described First chip is connect with phonetic incepting plate；Second chip is connect with action control plate.

Further, the holder,head includes facial frame, support plate and neck connector；The face frame is Empty hemisphere；There are two eyes through-hole, the blink mechanisms to pass through two for face frame surface Prosthetic Hand portion position processing Eyes through-hole is fixedly mounted on holder,head；Facial frame surface apery mouth position is machined with a plurality of circumferentially spaced logical Hole, the loudspeaker setting are convenient for sound transmission in through hole back-end；It is logical that the position of facial frame surface nose is machined with a circle Hole, Mike's external peripheral surface form interference with the through-hole inner surface and connect, the fixation for Mike；The neck connector It is arranged in neck luffing mechanism upper end.

Further, the blink mechanism includes blink steering engine, steering engine bracket, steering wheel, eyelid, eyeball, eyeball pillar；Institute It is affixed with support plate to state eyeball pillar；The eyeball and eyeball pillar are affixed；The eyelid is arranged in above eyeball, same with eyeball The heart and surface have certain gap not form friction, and one end of the eyelid and the output end of steering wheel connect；The input of the steering wheel It holds affixed with the output shaft of blink steering engine；The steering engine bracket is used to blink the affixed of steering engine and facial frame inner surface, described P2.3, VCC, the GND pin of second chip are connect with OUT, VCC, the GND pin of blink steering engine respectively.

Further, the voice receiver sheet and action control plate are arranged in the upper surface of support plate；The horn bracket Fixation for loudspeaker rear end face and lower end face of support plate.

Further, first chip and phonetic incepting intralamellar part speech chip press spi bus protocol communication, the first core PA4, PA5, PA6, PA7 pin of piece are connect with CS, SCK, MISO, MOSI pin of speech chip respectively, first chip PB15, PB12, PB13 pin connect respectively with RST, IRQ, WR pin in speech chip, the 5V of speech chip anode Voltage input end is connect with the output pin Vout of the positive low dropout voltage regulator of phonetic incepting intralamellar part, the speech chip SCK pin is connect with the PA5 pin of the first chip；The PA5 of first chip and the CLK pin of external crystal-controlled oscillation connect, and mention for system For clock pulses, TXD, RXD pin of first chip respectively at the main Bluetooth chip of phonetic incepting intralamellar part RXD, TXD pin connection, make phonetic incepting plate as with action control plate wireless communication in host, the positive and negative anodes pin of the loudspeaker Connected respectively at SPOP, SPON pin of speech chip, the positive and negative anodes pin of the Mike respectively at speech chip MICP, The connection of MICN pin.

Further, the neck swing mechanism includes first motor bracket, the one or four line two-phase stepper motor, back rotation Make auxiliary frame, neck pedestal；The output shaft of one or the four line two-phase stepper motor and revolution movement auxiliary frame are affixed, for passing Torque delivery；The revolution movement auxiliary frame is nested in the inner hole of neck pedestal；Under one or the four line two-phase stepper motor Surface and first motor bracket are affixed.

Further, P1.0, P1.1, P1.2, P1.3 pin of second chip are respectively in action control plate first B-IB, B-IA, A-IB, A-IA pin of motor control module connect；OUT1, OUT2 of first motor control module, OUT3, OUT4 pin is connect with 1,2,3,4 pins of the one or four line two-phase stepper motor respectively.

Further, the neck luffing mechanism is dynamic including the second electric machine support, the two or four line two-phase stepper motor, pitching Make auxiliary frame；The pitching motion auxiliary frame is nested in first motor bracket inside groove fixed；Two or the four line two-phase stepping The output shaft of motor is nested in affixed in the inner hole of pitching motion auxiliary frame；Second electric machine support is arranged in the two or four line two Phase step motor upper surface, and it is affixed therewith；Second electric machine support upper surface and neck connector lower end surface are affixed.

Further, P0.3, P0.2, P0.1, P0.0 pin of second chip are respectively in action control plate second B-IB, B-IA, A-IB, A-IA pin of motor control module connect；OUT1, OUT2 of second motor control module, OUT3, OUT4 pin is connect with 1,2,3,4 pins of the two or four line two-phase stepper motor respectively, the first motor control module and VCC, the GND pin of two motor control modules are connected respectively at the positive and negative anodes of 12V power supply, and TXD, RXD of second chip draw Foot respectively at RXD, TXD pin interconnection of the slave Bluetooth chip of action control intralamellar part, make action control plate as with language Slave in sound receiver board wireless communication.

A kind of control method realized based on the speech robot department of human head and neck device for the inquiry of factories and miness hospital department, The following steps are included:

Step a, Mike acquires voice messaging of the clients with noise as input signal in real time；

Step b, preliminary filtering processing is carried out to input signal using power spectrum subtraction method；

Step c, denoising is carried out using second-order EKF algorithm to the voice signal after preliminary filtering processing；

Step d, spike burr is eliminated using second order spreading kalman smoothing algorithm to step c treated voice signal；

Step e, filtered voice signal is input on the AD pin of speech chip；

Step f, speech chip initialization LD_Init_ASR () is written 0x21 to 0xb2 register, speech chip is made to enter sky Not busy state；

Step g, key words recognized list LD_AsrAddFixed () is added, wakes up chip to 0x37 register write-in 0x04 Entry work is added, recognizing addition entry when 0xbf register value is 0x31 terminates；

Step h, configuration 0x37 register is 0x06, wakes up chip and starts speech recognition work；

Step i, configuration 0x1c register opens converter, and wherein the value of register 0xdf register is 0x32 in identification process Between~0x3a, the mark of end of identification is that the value of 0xbf register is 0x35；

Step j, read register 0x2b value be 0x10, register 0xb2 value is 0x21, register 0xbf value be 0x35 simultaneously at It is vertical then identify success, it otherwise reconfigures register and prepares identification next time；

Step k, enter interrupt function Processint0 () when speech chip successfully recognizes audio signal, read 0xba and post Whether the value of storage is between 0~4, final to read in 0xc5 register if value illustrates there is correct recognition result between 0~4 The value in face obtains the correct answer of most probable, when the sound-playing function for recognizing correct voice and need to trigger speech chip；

Step l, the register LD_Init_MP3 () of first initial configuration MP3 corresponding function, by modification 0x8E register come Broadcast sound volume is adjusted, next configuration plays the related register of sound, and the third position of 0x1b register is set as 1, sequentially will Register 0x01 is written in MP3 data, and the data of MP3 are read from being stored in the first built-in chip type Flash；

Step m, MP3 data are sent into after speech chip, configure 0xba, 0x17 register, so that it may draw from SPOP, SPON Sound is exported in the loudspeaker of foot connection, realizes the function of voice road navigation.

Step n, simultaneously phonetic incepting plate is wirelessly communicated by main Bluetooth chip and action control plate shape journey, passes through string Mouthful send ASCII character, second chip according to different ASCII character control blink steering engines, the one or four line two-phase stepper motor, The completion of two or four line two-phase stepper motor blinks, nods, these interactivities of shaking the head act.

The present invention has the advantages that compared with the existing technology

1, the present invention provides a kind of speech robot department of human head and neck device for the inquiry of factories and miness hospital department and control, not only may be used To realize basic welcome's function, moreover it is possible to complete the work of voice road navigation, and can be made while voice road navigation Apery blink and a series of cervical motions solve the interactive difference of guest-meeting robot in the prior art and do not have department's inquiry function The problem of energy；It is directed to the factories and miness hospital that scale is smaller, medical worker is in short supply, the present invention solves factories and miness hospital for hospital guide side The personnel in face put into, and improve the overall image of hospital；

2, most of voice activity detection judgement, which is all based on ambient noise, (is swashed based on voice on stable this hypothesis VAD is surveyed in biopsy), due to registering in factories and miness hospital services hall, the less clients to stand in the queue to register of window are more, voice environment ten Point noisy, noise has the characteristics that the irregular pause of random non-stationary and voice, judges sound section according to the VAD of routine Unvoiced segments are difficult to find one or series of values, as the threshold value for judging sound section He unvoiced segments, so that this sentence Disconnected method becomes complicated, also results in influence to the estimation of noise, under the background noise environment of factories and miness hospital complexity, this hair Bright to use expanded Kalman filtration algorithm, providing a kind of voice signal has preferable constringent voice filter method, enhances language The accuracy of sound identification, improves the sensitivity of identification.

Detailed description of the invention

Fig. 1 is apparatus of the present invention structure chart；

Fig. 2 is apparatus of the present invention rear structure figure；

Fig. 3 is neck structure drawing of device of the present invention；

Fig. 4 is the first chip circuit of control system figure of the present invention；

Fig. 5 be speech chip of the present invention, main Bluetooth chip, external crystal-controlled oscillation and positive low dropout voltage regulator circuit diagram；

Fig. 6 is that second chip of the invention with from Bluetooth chip, steering engine of blinking connect circuit diagram；

Fig. 7 is first motor control module of the present invention and the second motor control module circuit diagram；

Fig. 8 is the filtering flow chart of Mike's input speech signal of the present invention；

Fig. 9 is the work flow diagram of speech chip of the present invention.

In figure: 1 robot head, 2 human-computer dialogue devices, 3 neck devices, 4 control systems, 1-1 holder,head, 1-2 blink Eye mechanism, 1-1A face frame, 1-1B support plate, 1-1C neck connector, 1-2A blink steering engine, 1-2B steering engine bracket, 1-2C Steering wheel, 1-2D eyelid, 1-2E eyeball, 1-2F eyeball pillar, 2-1 Mike, 2-2 phonetic incepting plate, 2-3 horn bracket, 2-4 loudspeaker , 2-5 action control plate, 3-1 neck swing mechanism, 3-2 neck luffing mechanism, 3-1A first motor bracket, 3-1B the one or four Line two-phase stepper motor, 3-1C revolution act auxiliary stand, 3-1D neck pedestal, the second electric machine support of 3-2A, 3-2B the two or four Line two-phase stepper motor, 3-2C pitching motion auxiliary frame.

Specific embodiment

Below with reference to attached drawing, the present invention is described in detail.

Specific embodiment one

For the speech robot department of human head and neck device of factories and miness hospital department inquiry, as shown in Figure 1 and Figure 2, including robot head 1, Human-computer dialogue device 2, neck device 3 and control system 4；The robot head 1 is arranged in 2 upper end of neck device, the people Machine Interface 3 is embedded in robot head 1；The robot head 1 includes holder,head 1-1 and blink mechanism 1-2；

The human-computer dialogue device 2 includes Mike 2-1, phonetic incepting plate 2-2, horn bracket 2-3, loudspeaker 2-4 and movement control Making sheet 2-5；

The neck device 3 includes neck swing mechanism 3-1 and neck luffing mechanism 3-2；The neck elevation mount 3 is arranged In 1 lower end of robot head；

The control system 4 is arranged in human-computer dialogue device 2；The control system 4 includes the first chip and the second chip；Institute The first chip is stated to connect with phonetic incepting plate 2-2；Second chip is connect with action control plate 2-5.

Specific embodiment two

As shown in Figure 1 and Figure 2, speech robot number of people on the basis of specific embodiment one, for the inquiry of factories and miness hospital department Neck device includes robot head 1, human-computer dialogue device 2, neck device 3 and control system 4, and the robot head 1 wraps It includes holder,head 1-1 and blink mechanism 1-2, the holder,head 1-1 includes that facial frame 1-1A, support plate 1-1B and neck connect Fitting 1-1C, the blink mechanism 1-2 include blink steering engine 1-2A, steering engine bracket 1-2B, steering wheel 1-2C, eyelid 1-2D, eyeball 1-2E, eyeball pillar 1-2F, the eyeball 1-2E and eyeball pillar 1-2F are affixed, and eyeball pillar 1-2F and support plate 1-1B are solid It connects, the eyelid 1-2D is arranged in above eyeball 1-2E, and surface concentric with the eyeball 1-2E has certain gap not formed to rub It wipes, the eyelid 1-2D and steering wheel 1-2C are affixed, the output shaft at the blink both ends steering engine 1-2A and the other end of steering wheel 1-2C Affixed, blink steering engine 1-2A is affixed by steering engine bracket 1-2B and the facial frame 1-1A, the neck connector 1-1C and Neck device 3 is affixed.

The human-computer dialogue device 2 includes Mike 2-1, phonetic incepting plate 2-2, horn bracket 2-3, loudspeaker 2-4 and moves Make control panel 2-5, the phonetic incepting plate 2-2 and action control plate 2-5 to be arranged on the support plate 1-1B, the Mike 2- 1 is fixed on the position of facial frame 1-1A apery nose, and the loudspeaker 2-4 is fixed on support plate 1-1B by horn bracket 2-3 On lower end surface.

The steering engine 1-2A that blinks in present embodiment is the prior art, the happy unreal limited public affairs of Sol electronics technology in manufacturer Shenzhen Department, model LDX227.

Mike 2-1 is the prior art, Nanjing Zhen Hong Science and Technology Ltd., manufacturer, model in present embodiment HM-700。

Loudspeaker 2-4 is the prior art, Ai Ning lighting Co., Ltd, manufacturer, model KS- in present embodiment 805。

As shown in Figure 2 and Figure 3, the neck device 3 includes neck swing mechanism 3-1 and neck luffing mechanism 3-2, described Neck swing mechanism 3-1 includes first motor bracket 3-1A, the one or four line two-phase stepper motor 3-1B, revolution movement auxiliary frame 3- 1C, neck pedestal 3-1D, the neck luffing mechanism 3-2 include the second electric machine support 3-2A, the two or four line two-phase stepper motor 3-2B, pitching motion auxiliary frame 3-2C；The output shaft of one or the four line two-phase stepper motor 3-1B and revolution act auxiliary frame 3-1C is affixed, and for transmitting torque, the revolution movement auxiliary frame 3-1C is nested in the inner hole of neck pedestal 3-1D, described to bow It faces upward movement auxiliary frame 3-2C and is nested in first motor bracket 3-1A inside groove fixed, the one or the four line two-phase stepper motor 3- The lower end surface 1B and the upper surface first motor bracket 3-1A are affixed, and the pitching motion auxiliary frame 3-2C and the two or four line two-phase walk Output axis connection into motor 3-2B passes through the second motor for transmitting torque, the two or the four line two-phase stepper motor 3-2B Bracket 3-2A is fixed on affixed with the neck connector 1-1C.

The one or four line two-phase stepper motor 3-1B is the prior art in present embodiment, and manufacturer's Zhejiang Jiaxing motor has Limit company, model 50BYJ.

The two or four line two-phase stepper motor 3-2B is the prior art in present embodiment, and manufacturer's Zhejiang Jiaxing motor has Limit company, model 50BYJ.

As shown in Figures 4 to 7, the control system 4 includes the first chip and the second chip, first chip and voice Receiver board 2-2 connection；Second chip is connect with action control plate 2-5, phonetic incepting plate 2-2 internal voice chip and first Chip with press spi bus protocol communication, PA4, PA5, PA6, PA7 pin of the first chip respectively with CS, SCK of speech chip, MISO, MOSI pin connection, PB15, PB12, PB13 pin of first chip respectively with RST, IRQ, WR in speech chip Pin connection, the 5V cathode voltage input terminal of the phonetic incepting plate 2-2 are connect with the anode of capacitor C1, C4, GND and capacitor The cathode of C1, C4 are connected, and are respectively connected to the Vin pin and ADJ/GND pin of positive low dropout voltage regulator, positive low pressure drop is steady The output pin Vout of depressor connect with capacitor C2, C3 export 3.3V voltage give entire circuit power supply, the SCK of the speech chip Pin is connect with the PA5 pin of the first chip；The PA5 of first chip and the CLK pin of external crystal-controlled oscillation connect, when providing for system Clock, RXD, TXD of main Bluetooth chip of TXD, RXD pin of first chip inside phonetic incepting plate 2-2 Pin connection, make phonetic incepting plate 2-2 as with action control plate 2-5 wireless communication in host, the loudspeaker 2-4 is just Negative pin is connected respectively at SPOP, SPON pin of speech chip, and the positive and negative anodes pin of the Mike 2-1 is respectively at voice core MICP, MICN pin of piece connects, P1.0, P1.1, P1.2, P1.3 pin of second chip respectively with action control plate 2- B-IB, B-IA, A-IB, A-IA pin connection of first motor control module in 5；OUT1, OUT2 of first motor control module, OUT3, OUT4 pin are connect with 1,2,3,4 pins of the one or four line two-phase stepper motor 3-1B respectively, second chip P0.3, P0.2, P0.1, P0.0 pin respectively with B-IB, B-IA of the second motor control module in action control plate 2-5, A-IB, The connection of A-IA pin；OUT1, OUT2, OUT3, OUT4 pin of second motor control module respectively with the two or four line two-phase stepping 1,2,3,4 pins of motor 3-2B connect, VCC, the GND pin of the first motor control module and the second motor control module It is connected respectively at the positive and negative anodes of 12V power supply, P2.3, VCC of second chip, GND pin are respectively with blink steering engine 1-2A's OUT, VCC, GND pin connection, slave bluetooth of TXD, RXD pin of second chip inside action control plate 2-5 RXD, TXD pin of chip connects, make action control plate 2-5 as with the slave in phonetic incepting plate 2-2 wireless communication.

The model STM32F103ZET6 of first chip；The model AT89C52 of second chip；The model of speech chip For LD3320；The model HC-08 of main Bluetooth chip；From the model HC-08 of Bluetooth chip；The type of positive low dropout voltage regulator Number be LM1117-3.3；The model HG7881 of first motor control module.

The course of work: the control system 4 includes the first chip and the second chip, first chip and phonetic incepting plate 2-2 connection；Second chip is connect with action control plate 2-5, voice road navigation function may be implemented, and in voice road Road can make apery blink and a series of cervical motions while guide.

Complete the course of work of voice road navigation: when the position of certain department in clients inquiry factories and miness hospital, voice Information is inputted by Mike 2-1, is carried out preliminary filtering processing to input signal by power spectrum subtraction method, is then expanded with second order It opens up Kalman filtering algorithm and carries out denoising, the voice signal after denoising is disappeared using second order spreading kalman smoothing algorithm Deglitch burr, voice signal is input to again on the AD pin of speech chip later, and analog signal is converted into number by AD pin Signal, by signer-independent sign language recognition technology (ASR) realize keyword identification extract, i.e., recognition result with deposit in advance The keyword message stored up in the first built-in chip type Flash compares, and identifies successfully and triggering is interrupted, execute voice road and refer to Draw function.

A series of course of work of apery blink and cervical motions: phonetic incepting plate 2-2 recognizes the language of interactive movement By Bluetooth wireless communication to action control plate 2-5 serially printing ASCII character after message breath, wherein main Bluetooth chip and from indigo plant Tooth chip needs to set the holotype and slave pattern of bluetooth on PC host computer in advance, and Setting pattern needs to turn by USB TTL electricity Flat conversion module is connect with PC host computer, sends AT instruction on host computer COM Debug Assistant to modify every ginseng of bluetooth It is several, it has configured main Bluetooth chip and the first chip in the later phonetic incepting plate 2-2 of master-slave relationship and has been connected to host, controlled to movement Information is sent by serial ports from the slave that Bluetooth chip and the second chip are constituted in making sheet 2-5, receives the second core after ASCII character Piece is completed to act accordingly according to the corresponding instruction control neck device 3 of different ASCII characters, controls the one or the four line two-phase step Into motor 3-1B, the two or four line two-phase stepper motor 3-2B and blink steering engine 1-2A rotation, to realize in voice guidance road Apery blink and a series of cervical motions can be made while road.

Specific embodiment three

A kind of control method realized based on the speech robot department of human head and neck device for the inquiry of factories and miness hospital department, including Following steps:

Step a, Mike acquires voice messaging of the clients with noise as input signal in real time；

Step b, preliminary filtering processing is carried out to input signal using power spectrum subtraction method；

Step c, denoising is carried out using second-order EKF algorithm to the voice signal after preliminary filtering processing；

Step d, spike burr is eliminated using second order spreading kalman smoothing algorithm to step c treated voice signal；Step E, filtered voice signal is input on the AD pin of speech chip；

Step f, speech chip initialization LD_Init_ASR () is written 0x21 to 0xb2 register, speech chip is made to enter sky Not busy state；

Step g, key words recognized list LD_AsrAddFixed () is added, wakes up chip to 0x37 register write-in 0x04 Entry work is added, recognizing addition entry when 0xbf register value is 0x31 terminates；

Step h, configuration 0x37 register is 0x06, wakes up chip and starts speech recognition work；

Step i, configuration 0x1c register opens converter, and wherein the value of register 0xdf register is 0x32 in identification process Between~0x3a, the mark of end of identification is that the value of 0xbf register is 0x35；

Step j, read register 0x2b value be 0x10, register 0xb2 value is 0x21, register 0xbf value be 0x35 simultaneously at It is vertical then identify success, it otherwise reconfigures register and prepares identification next time；

Step k, enter interrupt function Processint0 () when speech chip successfully recognizes audio signal, read 0xba and post Whether the value of storage is between 0~4, final to read in 0xc5 register if value illustrates there is correct recognition result between 0~4 The value in face obtains the correct answer of most probable, when the sound-playing function for recognizing correct voice and need to trigger speech chip；

Step l, the register LD_Init_MP3 () of first initial configuration MP3 corresponding function, by modification 0x8E register come Broadcast sound volume is adjusted, next configuration plays the related register of sound, and the third position of 0x1b register is set as 1, sequentially will Register 0x01 is written in MP3 data, and the data of MP3 are read from being stored in the first built-in chip type Flash；

Step m, MP3 data are sent into after speech chip, configure 0xba, 0x17 register, so that it may draw from SPOP, SPON Sound is exported in the loudspeaker of foot connection, realizes the function of voice road navigation.

Step n, phonetic incepting plate 2-2 and action control plate 2-5 sends ASCII character by bluetooth and forms channel radio to serial ports Letter, second chip is according to different ASCII character control blink steering engine 1-2A, the one or four line two-phase stepper motor 3-1B, the Two or four line two-phase stepper motor 3-2B completion blinks, nods, these interactivities of shaking the head act.

Claims (10)

1. for the speech robot department of human head and neck device of factories and miness hospital department inquiry, as shown in Figure 1 and Figure 2, including robot head (1), human-computer dialogue device (2), neck device (3) and control system (4)；The robot head (1) is arranged in neck device (2) upper end, the human-computer dialogue device (3) are embedded in robot head (1)；The robot head (1) includes holder,head (1- 1) and mechanism (1-2) is blinked；

The human-computer dialogue device (2) includes Mike (2-1), phonetic incepting plate (2-2), horn bracket (2-3), loudspeaker (2-4) And action control plate (2-5)；

The neck device (3) includes neck swing mechanism (3-1) and neck luffing mechanism (3-2)；The neck elevation mount (3) it is arranged in robot head (1) lower end；

Control system (4) setting is in human-computer dialogue device (2)；The control system (4) includes the first chip and second Chip；First chip is connect with phonetic incepting plate (2-2)；Second chip is connect with action control plate (2-5).

2. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 1, which is characterized in that The holder,head (1-1) includes facial frame (1-1A), support plate (1-1B) and neck connector (1-1C)；The face frame Frame (1-1A) is hollow hemisphere body；There are two eyes through-hole, institutes for face frame (1-1A) surface apery eye locations processing Blink mechanism (1-2) is stated to be fixedly mounted on holder,head (1-1) by two eyes through-holes；The facial frame surface (1-1A) is imitative People's mouth position is machined with a plurality of circumferentially spaced through-hole, and loudspeaker (2-4) setting is convenient for sound transmission in through hole back-end； The position of facial frame (1-1A) surface nose is machined with a circular through hole, and Mike's (2-1) external peripheral surface is logical with this Internal surface of hole forms interference connection, is used for the fixation of Mike (2-1)；The neck connector (1-1C) is arranged in neck pitching machine The upper end structure (3-2).

3. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 1, which is characterized in that The blink mechanism (1-2) includes blink steering engine (1-2A), steering engine bracket (1-2B), steering wheel (1-2C), eyelid (1-2D), eyeball (1-2E), eyeball pillar (1-2F)；The eyeball pillar (1-2F) and support plate (1-1B) are affixed；The eyeball (1-2E) and eye Ball pillar (1-2F) is affixed；The eyelid (1-2D) is arranged in above eyeball (1-2E), concentric with eyeball (1-2E) and surface has Certain gap does not form friction, and one end of the eyelid (1-2D) is connect with the output end of steering wheel (1-2C)；Steering wheel (the 1- Input terminal 2C) and the output shaft of blink steering engine (1-2A) are affixed；The steering engine bracket (1-2B) is for blinking steering engine (1-2A) It is affixed with facial frame (1-1A) inner surface, P2.3, VCC of second chip, GND pin respectively with blink steering engine (1- OUT, VCC 2A), GND pin connection.

4. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 1, which is characterized in that The voice receiver sheet (2-2) and action control plate (2-5) are arranged in the upper surface of support plate (1-1B)；The horn bracket (2-3) is used for the fixation of loudspeaker (2-4) rear end face and the lower end surface support plate (1-1B).

5. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 4, which is characterized in that First chip and phonetic incepting plate (2-2) internal voice chip press spi bus protocol communication, PA4, PA5 of the first chip, PA6, PA7 pin are connect with CS, SCK, MISO, MOSI pin of speech chip respectively, PB15, PB12 of first chip, PB13 pin is connect with RST, IRQ, WR pin in speech chip respectively, the 5V cathode voltage input terminal and language of the speech chip The output pin Vout connection of the internal positive low dropout voltage regulator of sound receiver board (2-2), the SCK pin of the speech chip and The PA5 pin of first chip connects；The PA5 of first chip and the CLK pin of external crystal-controlled oscillation connect, and provide clock arteries and veins for system TXD, RXD pin of punching, first chip draw respectively at RXD, TXD of the internal main Bluetooth chip of phonetic incepting plate (2-2) Foot connection, make phonetic incepting plate (2-2) as with action control plate (2-5) wireless communication in host, the loudspeaker (2-4) Positive and negative anodes pin is connected respectively at SPOP, SPON pin of speech chip, and the positive and negative anodes pin of the Mike (2-1) is respectively at language MICP, MICN pin of sound chip connect.

6. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 1, which is characterized in that The neck swing mechanism (3-1) includes first motor bracket (3-1A), the one or four line two-phase stepper motor (3-1B), back rotation Make auxiliary frame (3-1C), neck pedestal (3-1D)；The output shaft and back rotation of one or the four line two-phase stepper motor (3-1B) It is affixed to make auxiliary frame (3-1C), for transmitting torque；Revolution movement auxiliary frame (3-1C) is nested in neck pedestal (3-1D) Inner hole in；The lower surface of one or the four line two-phase stepper motor (3-1B) and first motor bracket (3-1A) are affixed.

7. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 6, which is characterized in that P1.0, P1.1, P1.2, P1.3 pin of second chip respectively with first motor control module in action control plate (2-5) B-IB, B-IA, A-IB, A-IA pin connection；OUT1, OUT2, OUT3, OUT4 pin of first motor control module respectively with 1,2,3,4 pins of the one or four line two-phase stepper motor (3-1B) connect.

8. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 1, which is characterized in that The neck luffing mechanism (3-2) is dynamic including the second electric machine support (3-2A), the two or four line two-phase stepper motor (3-2B), pitching Make auxiliary frame (3-2C)；The pitching motion auxiliary frame (3-2C) is nested in first motor bracket (3-1A) inside groove fixed；Institute State the two or four line two-phase stepper motor (3-2B) output shaft be nested in it is affixed in the inner hole of pitching motion auxiliary frame (3-2C)；Institute The second electric machine support (3-2A) setting is stated in the two or four upper surface line two-phase stepper motor (3-2B), and affixed therewith；Described The two electric machine supports upper surface (3-2A) and the lower end surface neck connector (1-1C) are affixed.

9. the speech robot department of human head and neck device for the inquiry of factories and miness hospital department according to claim 8, which is characterized in that P0.3, P0.2, P0.1, P0.0 pin of second chip respectively with the second motor control module in action control plate (2-5) B-IB, B-IA, A-IB, A-IA pin connection；OUT1, OUT2, OUT3, OUT4 pin of second motor control module respectively with 1,2,3,4 pins of the two or four line two-phase stepper motor (3-2B) connect, the first motor control module and the second motor control The VCC of molding block, GND pin are connected respectively at the positive and negative anodes of 12V power supply, TXD, RXD pin of second chip respectively at RXD, TXD pin interconnection of the internal slave Bluetooth chip of action control plate (2-5), make action control plate make (2-5) be with Slave in phonetic incepting plate (2-2) wireless communication.

10. a kind of realized based on any speech robot department of human head and neck device for the inquiry of factories and miness hospital department of claim 1 to 9 Control method, which comprises the following steps:

Step a, Mike (2-1) acquires voice messaging of the clients with noise as input signal in real time；

Step b, preliminary filtering processing is carried out to input signal using power spectrum subtraction method；

Step c, denoising is carried out using second-order EKF algorithm to the voice signal after preliminary filtering processing；

Step d, spike burr is eliminated using second order spreading kalman smoothing algorithm to step c treated voice signal；

Step e, filtered voice signal is input on the AD pin of speech chip；

Step f, speech chip initializes LD_Init_ASR(), 0x21 is written to 0xb2 register, speech chip is made to enter the free time State；

Step g, key words recognized list LD_AsrAddFixed(is added), chip, which is waken up, to 0x37 register write-in 0x04 adds Entry is added to work, recognizing addition entry when 0xbf register value is 0x31 terminates；

Step h, configuration 0x37 register is 0x06, wakes up chip and starts speech recognition work；

Step i, configuration 0x1c register opens converter, and wherein the value of register 0xdf register is 0x32 in identification process Between ~ 0x3a, the mark of end of identification is that the value of 0xbf register is 0x35；

Step j, read register 0x2b value be 0x10, register 0xb2 value is 0x21, register 0xbf value be 0x35 simultaneously at It is vertical then identify success, it otherwise reconfigures register and prepares identification next time；

Step k, enter interrupt function Processint0(when speech chip successfully recognizes audio signal), it reads 0xba and posts Whether the value of storage is between 0 ~ 4, final to read inside 0xc5 register if value illustrates there is correct recognition result between 0 ~ 4 Value, the correct answer of most probable is obtained, when the sound-playing function for recognizing correct voice and need to trigger speech chip；

Step l, the register LD_Init_MP3(of first initial configuration MP3 corresponding function), it is adjusted by modification 0x8E register Broadcast sound volume is saved, next configuration plays the related register of sound, the third position of 0x1b register is set as 1, sequentially by MP3 Register 0x01 is written in data, and the data of MP3 are read from being stored in the first built-in chip type Flash；

Step m, MP3 data be sent into speech chip after, configure 0xba, 0x17 register, so that it may from SPOP, SPON pin Sound is exported in the loudspeaker (2-4) of connection, realizes the function of voice road navigation；

Step n, simultaneously phonetic incepting plate (2-2) is wirelessly communicated by main Bluetooth chip and action control plate (2-5) shape journey, ASCII character is sent by serial ports, second chip is according to different ASCII character control blink steering engine (1-2A), the one or four line Two-phase stepper motor (3-1B), the completion of the two or four line two-phase stepper motor (3-2B) blink, nod, these interactivities of shaking the head are dynamic Make.