AVR-based speech recognition assistants help the disabled indoor walking service system design

I. Project Overview

1.1 Introduction

At present, the number of elderly people over the age of 60 in the world has reached 600 million, and the aging of the population has become a major social problem worldwide. China's elderly population aged 60 and over has reached 132 million, accounting for 10% of the country's total population, and continues to grow at an average annual rate of 3.32%. It is estimated that by 2015, the population over the age of 60 will exceed 200 million, accounting for about 14% of the total population, and about 50% of the elderly patients need nursing services. In addition, there are about 60 million disabled people in China, accounting for about 5% of China's total population. As a result, the total number of people in China who need care is about 100 million. The accelerated ageing of the population puts the medical and social insurance systems under unprecedented pressure. Among the 1.3 billion people in China, the number of families is about 300-400 million. The families with dysfunction or those who need caregivers account for 1/10 of the total number of households in the country, about 35.77 million. The large number of disabled people and long-term bedridden elderly people make the problem of helping the elderly and helping the disabled become a major social problem. Improving the quality of life and self-care ability of the elderly and the disabled, researching and developing products for helping the elderly and helping the disabled, and providing services such as walking and indoor living for the elderly and the disabled have become a common concern of the whole society.

1.2 Project Background / Motivation

At present, there are more than 30 million elderly and disabled people with physical disabilities in our country. When walking or elderly people with disabilities, they need to use human resources or equipment, such as wheelchairs. Most of the current home appliances do not support voice recognition. The switch of the appliance is usually implemented by a button or a remote control. Before the control is implemented, the precise position or direction of the controlled device (such as buttons, remote objects, etc.) must be found, which is very troublesome for the user to be inconvenient to move.

Therefore, it has important practical significance for the development of the old-age disability system. It plays an important role in improving the quality of life of the elderly and the disabled and ensuring the harmony, stability and development of our society.

In order to actively carry out the science and technology innovation practice activities of college students, it is also for the students to develop the learning ability of AVR single-chip microcomputer, so this question is selected to study AVR microcontroller.

Second, the demand analysis

2.1 Functional requirements

The voice control can effectively control the wheelchair to walk freely;

Realize automatic control of home appliances in the home;

Realize the control of the access control system;

According to the degree of knowledge mastery, the family security information can be monitored and released in real time.

2.2 Performance requirements

The system uses voice to issue related control commands and alarm commands;

The household appliances and their peripheral devices are all implemented by simulation functions;

Use a four-wheeled vehicle model instead of a wheelchair to move freely around the front, rear, left and right;

The electric lamp replaces the switch quantity of the home appliance, and realizes the switch control of the household appliance;

Simulate the automatic switching function of the access control using a DC motor;

Using the timer software of the AVR microcontroller to simulate the remote control to realize the remote control function of the household appliance;

The function of dialing and SMS alarm of mobile phone is realized by GPRS (depending on the ability).

Third, the program design

3.1 System function realization principle

The overall design scheme of the AVR-based voice recognition helper and disabled indoor walking service system is shown in Figure 1. The voice signal is collected by the voice processing module through the microphone on the development board, and then sent to the AVR processor for identification. The AVR microprocessor analyzes and determines the type of the voice according to the recognition result, and completes the corresponding operation according to the instruction requirement. The AVR microprocessor sends the related operations to the corresponding execution device such as the access control device, the wheelchair drive device, etc. through the wireless module, accepts the command and completes the corresponding action; the system can monitor the outdoor situation through the camera, and the video information It is transmitted to the LCD display for real-time display. The user decides whether to open the access control system according to the identity of the visitor; realizes automatic dialing of the mobile phone through voice, and uses the mobile phone to make calls and help.

Figure 1 Overall block diagram of the system

3.2 Hardware platform selection and resource configuration

The hardware platform selects an EVK1100 development system based on AVR32 AT32UC3A microcontroller controller. The specific resource configuration is as follows:

The basic input and output ports of the AT32UC3A MCU are used to control the switching quantity;

The serial communication interface of the single chip microcomputer is used for communication of the voice module and the wireless control module;

The PWM of the single chip microcomputer is used to control the steering and speed control of the wheelchair;

The Ethernet port is used for the video input of the monitor;

Development system sensors (light, temperature) for humanized prompts and control of household appliances, weather forecasts and other functions;

The 4x20 blue LCD (PWM variable backlight) is used to display control information and its associated status.

3.3 system software architecture

The system software environment uses the official development tool AVR Studio 5 provided by ATMEL, and the program is mainly programmed in C language.

3.4 System Software Process

Main program design

Figure 2 main program flow chart

In the main program, the modular design is used to complete the training and storage of the voice of a specific person. After the training is successful, the voice recognition is performed, and the related operations are performed according to the recognized commands, as shown in FIG. 2 .

Speech recognition software design