I. Introduction
1. 1 infrared communication technology [8]
In the early 1960s, some developed countries began to study the remote control technology of civilian products, but the development of remote control technology was very slow due to the technical conditions at that time. In the late 1970s, with the development of large-scale integrated circuits and computer technology, remote control technology developed rapidly. The remote control mode has gone through several stages: from wired to wireless ultrasonic, from vibrator to infrared, and then to microcomputer infrared remote control using bus. No matter which method is adopted, it is very important to transmit the signal accurately and finally achieve satisfactory control effect. The original wireless remote control device used electromagnetic waves to transmit signals. Because electromagnetic waves are easily disturbed, they are gradually used to transmit signals. Compared with infrared ray, ultrasonic sensor has a narrow frequency band, can carry less information, and is easily disturbed, resulting in misoperation. The most ideal way is light control, and infrared remote control has gradually replaced ultrasonic remote control, and infrared multifunctional remote control has appeared, which has become the mainstream of today's times. Because infrared is outside the visible light spectrum, it has strong anti-interference and linear propagation characteristics of light waves, and it is not easy to interfere with each other, so it is a good information transmission medium. Information can be transmitted by directly modulating infrared light, for example, information can be transmitted by directly modulating the intensity of infrared light, or it can be used to generate a carrier with a certain frequency, and then the carrier is modulated by information, and then the receiver removes the carrier to obtain information. From the reliable transmission of information, the latter method is better, which is also the method adopted by most infrared remote controllers we see today. In recent ten years, infrared remote control technology has developed rapidly, especially in the fields of color TV, DVD, air conditioning and other household appliances. , has also been widely used in other electronic fields. With the improvement of people's living standards, the pursuit of products is more convenient and intelligent, and infrared remote control technology is a key development direction.
Infrared is an electromagnetic wave with a wavelength between 750nm and 1mm, and its frequency is higher than that of microwave and lower than that of visible light, which is invisible to human eyes. Infrared communication generally uses near infrared in infrared band, and the wavelength is between 0.75um-25um and 25 um. At present, almost all the widely used home appliance remote controllers adopt infrared transmission technology. As the transmission mode of wireless local area network, the biggest advantage of infrared mode is that it is not interfered by radio and its use is not restricted by the National Radio Regulatory Commission. After the establishment of the Infrared Data Association (IRDA), in order to ensure that infrared products from different manufacturers can obtain the best communication effect, the infrared communication protocol limits the wavelength range of light waves used in infrared data communication to 850-900nm.
1.2 speech signal recognition technology
Speech recognition is a high-tech machine, which converts speech signals into corresponding text files or commands through the process of recognition and understanding. As a special research field, speech recognition is an interdisciplinary subject, which is closely related to acoustics, phonetics, linguistics, digital signal processing theory, information theory, computer science and many other disciplines. After more than forty years of development, speech recognition has shown great application prospects. To some extent, it is similar to the difficulty for a machine to recognize speech that a person with poor foreign language listens to a foreigner. It is related to different speakers, different speech speeds, different speech contents and different environmental conditions. The characteristics of speech signal itself bring difficulties to speech recognition. These characteristics include variability, dynamics, instantaneity and continuity.
The research of speech recognition began in 1950s, when AT & amp； T Bell Laboratories has realized the first speech recognition system which can recognize ten English numbers-AUDRY system.
In 1960s, the application of computers promoted the development of speech recognition. The important achievement of this period is the proposal of dynamic programming (DP) and linear predictive analysis (LP), in which the latter solves the problem of speech signal generation model and has a far-reaching impact on the development of speech recognition.
In 1970s, a breakthrough was made in the field of speech recognition. Theoretically, LP technology is further developed, and dynamic time normalization technology (DTW) is basically mature, especially the theory of Vector Quantization (VQ) and Hidden Markov Model (HMM). In practice, a speaker-specific isolated speech recognition system based on linear predictive cepstrum and DTW technology is realized.
In 1980s, the research of speech recognition was further deepened, and its notable feature was the successful application of HMM model and artificial neural network (ANN) in speech recognition. The wide application of HMM model should be attributed to the efforts of Rabiner of Bell Laboratories of AT&T and other scientists, who have engineered the originally difficult HMM pure mathematical model so that more researchers can understand and understand it. The speech recognition system established by ANN and HMM models has the same performance.
In the 1990s, with the advent of the multimedia era, it is urgent for the speech recognition system to move from laboratory to practical use. Many developed countries, such as America, Japan, Korea, IBM, Apple, AT&T, NTT and other famous companies, have invested heavily in the practical development and research of speech recognition systems.
The research of speech recognition in China has been closely following the international level, and the country attaches great importance to it. The research of large vocabulary speech recognition is included in the "863" plan, which is jointly developed by Institute of Acoustics, Institute of Automation, Chinese Academy of Sciences and Peking University. In view of the huge market in China in the future, foreign countries also attach great importance to the study of Chinese speech recognition. The United States, Singapore and other places have gathered a group of scholars from Chinese mainland, Taiwan Province Province, Hongkong and other places, and their research achievements have reached a fairly high level. Therefore, in addition to strengthening theoretical research, it is also necessary to speed up the transformation of laboratory demonstration systems into commodities in China.
At present, the research progress of speech recognition is slow, mainly because there has been no breakthrough in theory. Although various new correction methods are emerging, their universal applicability is controversial. Specifically, the difficulties are mainly manifested in:
1. The poor adaptability of speech recognition system is mainly reflected in its strong dependence on the environment, that is, the speech training system collected in a certain environment can only be applied in this environment, otherwise the system performance will drop sharply; Another problem is that it can't correctly respond to the user's wrong input and is inconvenient to use.
2. It is difficult to make progress in speech recognition in high noise environment, because people's pronunciation changes greatly at this time, such as louder voice, slower speech speed, and changes in tone and vibration peak. This is the so-called Lombardy effect, and new signal analysis and processing methods must be found.
3. There have been many research achievements in linguistics, physiology and psychology, but how to quantify, model and apply these knowledge to speech recognition needs to be studied. Language model, grammar model and vocabulary model are very important in continuous speech recognition with large vocabulary.
4. Our understanding of human auditory understanding, knowledge accumulation and learning mechanism and the control mechanism of brain nervous system is still unclear; Secondly, it is still a difficult process to apply the existing achievements in this field to speech recognition.
5. There are still many specific problems to be solved in the transformation of speech recognition system from laboratory demonstration system to commodity, such as recognition speed, rejection and keyword (sentence) detection technology (that is, removing "ah" and "alas" from continuous speech to obtain the real speech part to be recognized) and other technical details to be solved.
There are many difficulties in speech recognition, but the key to effective recognition is its model and algorithm. Model algorithm is the key to whether the computer's ability to describe speech can grasp the essence of human speech. In the field of speech recognition, although there is competition between financial strength and human resources, the most fundamental thing is the competition between its key core technologies-models and algorithms.
1.3 the purpose and basic idea of this work design.
This work uses PC to realize the intelligent control of infrared control technology and voice signal recognition technology in household appliances or industrial electrical equipment, as well as the application in PC software systems, such as intelligent household appliances control system and multimedia teaching control system. In addition, the computer system or file can be encrypted effectively by using the encryption means of infrared ray and voice recognition technology.
With the development of society and the improvement of national economic level, more and more PCs have entered ordinary people's homes, industrial fields and school teaching fields. This work uses the powerful functions of the existing PC, combines infrared control technology and voice recognition technology, and realizes flexible control of household appliances or factory electrical equipment and lively multimedia teaching means. Under the condition of not occupying too much PC resources, use PC resources as much as possible, minimize the front-end hardware system and minimize the cost. In addition, the perfect combination of wireless control technology and infrared communication technology is adopted, which avoids the space limitation of simple infrared communication and the incompatibility between simple wireless communication and existing infrared electrical equipment, and makes the system control more flexible. In addition, the system can also realize remote control of equipment by connecting to the Internet.
1. Use infrared receiving circuit to identify the signals sent by various infrared remote controllers through USB port on PC, then analyze and store the signals, and finally output them to radio receiving and infrared transmitting controller through USB port to reproduce their infrared control signals, so as to control the equipment. As shown in figure 1- 1. In this process, PC plays the role of learning remote control. Compared with the learning remote controller developed by ordinary single chip microcomputer, this method can avoid the advantage of increasing hardware cost. In addition, the powerful processing function of PC is beyond the reach of ordinary learning remote controller.
Figure 1- 1 schematic diagram of equipment control
2. The voice signal of this work is input through MIC, and the voice signal recognition technology combined with the powerful processing function of PC can effectively recognize the voice signals of different people. After being processed, the voice signals are converted into infrared control signals or other effective signals to control or output PC application software control equipment.
3. The system is equipped with a PC-specific remote controller (as shown in figure 1-4), which can be used to control the PC software layer and external devices, equivalent to a learning remote controller. Using this special infrared remote controller, flexible multimedia teaching methods can be realized. As shown in figure 1-3.
Figure 1-3 multimedia teaching control chart 1-4 PC special remote controller
4, the traditional keyboard password input, Trojan can easily record the input key information, or record your key information through a micro camera. Cryptography technology based on infrared analysis and speech recognition technology can encrypt systems and files more conveniently and effectively.
1.4 Summary of the current research level of similar topics at home and abroad
There is no reference website for the product registration in the patent query database provided by China, China National Intellectual Property Administration and China Patent Information Network, and there is no registration of the product in the foreign search linked by China Patent Information Network (such as European Patent Office and US Patent Office).
1, infrared remote control technology is relatively mature, and learning remote control is a good example. It accurately records the waveform of the remote controller one by one at the receiver of the learning remote controller, and then reproduces and emits the waveform when it is applied, but it is limited by memory and the stored signal is limited. After patent search, it is found that there are infrared remote control PC products at home and abroad, and the Japanese Tsukiji company also applied for a patent for computer remote control in China. Most products use RS232 or PS2 port to control PC, but there is no product that controls household appliances based on USB interface.
2. There are many voice control products realized by hardware, such as intelligent toy robot dog, Japanese robot dog "Aibo" and British robot dog "Wake", all of which are electronic pets based on voice control, but they are expensive. At present, IBM, Apple, AT&T, NTT and other famous companies have invested heavily in the practical development and research of speech recognition systems. The main product is voice input method, and there is no flexible control method using voice in equipment control.
3. The home intelligent control system realized by pure hardware has a variety of traditions, mostly using infrared and wireless technologies, but voice control home products have not yet appeared. These home control products have no computer interface, so it is impossible to realize home management through computers.
4. Traditional user authentication is generally through keyboard, such as system encryption or file encryption. There is no computer authentication method through infrared analysis or speech recognition at home and abroad. Registered patents are password identification products, generally password locks and communication password authentication technologies.
1.5 work instruction
1.5. 1 hardware installation
Plug the MIC into the sound card of the PC, plug the infrared receiver and radio transmitter (Figure 1-5) into the USB interface of the PC, install the radio receiver and infrared transmitter controller (Figure 1-6) in the direction of the controlled equipment, and then connect the 220V power supply. When the controlled equipment has no infrared data association, the power plug of the controlled equipment is inserted into the module to realize the power control of the equipment.
Figure 1-5 infrared receiver and radio transmitter
Figure 1-6 Radio Receiving and Infrared Transmitting Controller
1.5.2 Software usage
Minimum software configuration requirements: memory 64M, Pentium 2 processor, WIN98/ME/2K/XP system.
After the software is installed, the control window as shown in figure 1-7 will appear and enter the setting bar. After selecting the type of remote controller, aim the remote controller of the controlled equipment at the infrared receiver and radio transmitter (Figure 1-5) to input infrared remote control signals, and voice signals can be input at the same time. After the operation is completed, the input of infrared signal and voice signal is completed, and other data can also be selected according to this method. This work can not only control TV, air conditioner and water heater, but also control PC itself. In addition, other electrical equipment can be customized. As shown in figure 1-8, select the network bar to control the remote device. The specific usage method is prompted in the software help, and the operation is very simple.
1.6 market analysis and forecast
1. The hardware cost of this product is low (below 200 yuan), and the USB device supports hot plug, which is convenient for users to install and use. The software system is connected with Microsoft Office tools, and can control the audio-visual education system and peripheral equipment through remote control and voice, such as the demonstration and switching of multimedia documents, the transformation of lighting and sound, etc.
2. The family group equipped with computers is huge, which contains huge business opportunities. Home users can use the entertainment control functions provided by this product, such as operating a voice or video player with a remote controller or voice, and controlling game software with a remote controller and voice. Of course, its smart home appliance control can make users realize the convenience brought by information technology. At present, with the rapid development of the network, most portable devices support network communication except PC, and it will be welcomed by everyone to use the network to remotely control electrical equipment.
3. The password system based on infrared analysis and voice recognition can prevent the internal information of the computer from being copied or changed by others, and the operation is easier than the traditional key password input. It can greatly improve the security and reliability of information and data confidentiality of individuals and enterprises.