KR20030028351A - Joystick and Voice Recognition Electric Wheelchair by Artificial Intelligence - Google Patents

Abstract

PURPOSE: Provided is an electric wheel chair for the physically handicapped person, recognizing voices and having a joystick using artificial intelligence. The wheel chair has enhanced stability, reliability and conveniency, recognizes the surrounding environment and controls the movements by using artificial intelligent control system. CONSTITUTION: The wheel chair comprises: an intelligent control module with hardware and software which controls the whole system; a joystick module, speech recognition module and environment perception module which receive users input data; a power module which supplies power, being controlled by the control module; and a display module which exhibits the operation state of the wheel chair.

Description

Joystick and Voice Recognition Electric Wheelchair by Artificial Intelligence}

The present invention is to recognize the surrounding environment by the sensor to improve the convenience, safety, reliability, etc. of the wheelchair widely used as a means of movement of the disabled and also to control the operation of the wheelchair optimally in such a dynamic environment by using an intelligent control technique. The present invention relates to a joystick and an electric wheelchair for voice recognition using artificial intelligence .

The technical status of the wheelchair system currently being commercialized or researched is as follows.

(1) Overseas Status

▶ In developed countries, which have long worked for the convenience of the disabled, various studies have been conducted to make the electric wheelchair more functional.

▶ Research is being conducted to introduce voice recognition technology or image processing technology through cameras for users who are unable to use the joystick, which is the driving device of electric wheelchairs, as well as general manual wheelchairs due to severe disability.

▶ Products that consider the safety of users are being developed to cope with changes in the surrounding environment through multiple sensors.

▶ In order to optimally provide the driving environment for the disabled, all parts used in the chair are packaged or modularized so that the user can select them according to their functions to provide customized services considering the characteristics of each user.

(2) Domestic status

▶ Currently, the use of wheelchairs by handicapped people is mainly dependent on passive wheelchairs, but the demand for electric wheelchairs is gradually increasing for the convenience and safety of the disabled.

▶ Due to the small size of domestic wheelchair manufacturers, the development of electric wheelchair technology is inferior to other advanced countries. Therefore, even though most companies sell or manufacture imported goods , many major parts such as motors, joysticks, and controllers, which are the main parts of electric wheel chairs , depend on imports .

▶ Actually, it is known that people with disabilities are in urgent need of such electric wheelchairs, but the use rate of electric wheelchairs by domestic disabled people is inevitably lower than that of other developed countries. It is not popularized.

▶ Contrary to the domestic situation, the domestic technology level for electric wheelchair development is enough to develop products or researched in various fields, so it is easier to introduce and commercialize these technologies in a short time than any other fields. It is considered to be.

* Electric wheelchair for joystick and voice recognition using artificial intelligence technique for convenience and safety when people with disabilities move

① AI controller : Program and control device using intelligent control technique to control the operation of electric wheelchair

② Environmental recognition system : Sensor equipment and interface technology using multiple sensors to recognize dynamic surroundings while driving

③ Speech Recognition System : General purpose speech recognition system using artificial intelligence technique that can drive driving command using few simple words besides joystick for convenience and usage of disabled people, and also has foreign language recognition ability for foreign users.

1 is a schematic diagram of the entire system.

The intelligent electric wheelchair system is composed of several modules as shown in the drawing.

In other words, Joystick Module that accepts user's input , Speech Recognition Module , Environment Perception Module that recognizes the surrounding environment during wheelchair operation, Intelligent Control Module that controls all systems, Real controller while supplying real power of motor It consists of a power module that is the control target of the control module and a display module that shows the driving status of the wheelchair.

[Figure 2]-Structure of the whole system

Fig. 3- Outline drawing of electric wheelchair for joystick and voice recognition using artificial intelligence technique

① Ultrasound Sensor

② Caster Wheel

③ Battery and Control Board

④ Rear Wheel

⑤ Joystick, Display Module, Microphone

⑥ Provide removable removable function

Figure 4- Structure of intelligent control module using 80C196KC

[Figure 5] -Structure of Intelligent Electric Wheelchair Control Algorithm

6 is a block diagram of input parameter calculation using a neural network.

Fig. 7- Fuzzy Reasoning System for Controlling Motor Speed

[Figure 8] -Joystick Structure

Figure 9- Rotational direction of the motor according to the joystick position

FIG. 10- Structure of Time Delay Neural Network (TDNN) for Speech Recognition

Figure 11- Sensor System

▶ intelligent controller capable of controlling effectively wheelchair system by introducing neural networks and advanced technologies such as fuzzy instead of artificial intelligent control technology, human learning and decision-making skills with intelligent implementation of the human capacity

▶ Environment recognition system using multiple sensors (ultrasound sensor) to proactively cope with the variable situation of the wheelchair environment during wheelchair operation

Intelligent voice recognition algorithm that operates a wheelchair with a few simple words for the convenience and usage of the disabled

▶ Wheelchair integrated control system to apply various developed algorithms and control wheelchair operation according to each input

1. Intelligent Control Module: See [Fig. 4]-[Fig. 7].

▶ Controller configuration using Microprocessor 80196KC

▶ Control input accepts joystick input or voice input according to user's selection

▶ As a control algorithm, fuzzy inference system, an intelligent control algorithm, is applied.

▶ As output, 80196KC PWM output is used to control the motor connected to the rear wheel of the wheelchair.

Intelligent Control

It is effective for systems whose political correlations between variables are not clearly identified but are somewhat empirically or qualitatively identified.

ㆍ It is effective when you want to establish the correlation of data by learning data by experiment and experience.

Applicable to systems where the system is applied under conditions of strong nonlinearity, uncertainty and ambiguity

2. Joystick Module: See [Fig. 8]-[Fig. 9].

▶ The direction and speed of rotation of the two motors that represent the wheelchair movement are determined by the user's joystick

▶ The fuzzy filter algorithm is used to increase the precision of the joystick input signals Vml and Vmr.

3. Speech Recognition Module: See FIG.

▶ Design voice recognition system using DSP Board using TMS320C31

▶ Speech recognition algorithm using TDNN (Time Delay Neural Network) that recognizes temporal features in the input pattern by using delay factors after preprocessing of voice input.

▶ The user manipulates the wheelchair with a combination of nine words

→ "Forward", "Back", "Left", "Right", "Slow", "Fast". "Keep", "Stop", "Rotate"

▶ Universal voice recognition function for English, Japanese, Chinese and Spanish for foreign users

4. Environment Perception Module: see FIG.

▶ Use 10 ultrasonic sensors to avoid obstacles not recognized by the user

→ Measure the direction and distance of obstacles

▶ Use the shaft encoder mounted on the motor to measure the speed of the wheelchair

→ The speed of the wheelchair can be calculated by measuring the speed of the motor.

The expected effects due to the present invention are as follows.

▶ Fuzzy-neuro algorithm, an intelligent control technique, can be used to optimally control wheelchair movement in a dynamic environment, which is expected to improve safety of wheelchair users.

▶ The function to control the input of the electric wheelchair by voice other than the joystick is added to the system to maximize the range of use according to the degree of disability, and also for foreign users other languages such as English, Japanese, Chinese, Spanish, etc. It is possible to recognize the expansion of the export market of the present invention is possible.

▶ For the safety of users, the environmental recognition module using multiple sensors can recognize invisible obstacles or dynamic obstacles, and together with the intelligent control module, there is a great expectation for improving the safety and reliability of wheelchair operation for the disabled.

▶ Also, due to the small size of domestic companies, import dependence is higher than technology development, but in the present invention, it is possible to achieve import substitution effect by developing and manufacturing technical parts such as controllers, and it is very effective for improving national competitiveness and technological power. It is considered to be.

Claims (1)

The joystick and voice recognition electric wheelchair using the artificial intelligence technique of the present invention implements an artificial intelligence electric wheelchair controller using the Fuzzy-neuro algorithm, which is an artificial intelligence algorithm, as an implementation algorithm of the controller for enhancing safety and convenience. In addition, the controller implements 80c196kc and accepts a joystick or voice input as the user's choice, and uses the PWM output for motor control of the electric wheelchair by the artificial intelligence control algorithm. In addition, the speech recognition system can be implemented using TMS320C31 for speech recognition that can maximize the range of use according to the degree of failure, and TDNN (Time Delay Neural Network) is used as the intelligent speech recognition algorithm. In addition, TDNN will be taught to recognize foreign languages such as English, Japanese, Chinese, and Spanish for foreign users. Wheelchair users can use simple speech combinations ("forward", "back", "left", "right", "slow", "fast", "keep", "stop" "Rotate" to operate the wheelchair. In order to avoid the obstacles that the wheelchair user can recognize or to avoid moving obstacles, 10 ultrasonic sensors are used to measure the direction and distance of the obstacle, and the current direction and speed of the wheelchair are measured to increase the risk of collision. It is equipped with a function to increase the safety of users by constructing an environmental recognition module that prevents collision by the electronic brake system of the motor.