Design of telerehabilitation information collection system based on fuzzy control

1 Introduction

　　Telerehabilitation is a multidisciplinary cross-disciplinary subject that combines modern information and communication technology with rehabilitation medicine. It can be defined as: remote rehabilitation medical services achieved on the basis of comprehensive use of communication, remote sensing, remote control, computers, information processing and other technologies.

　　Foreign research in this area has different starting points. In summary, it mainly regards the telerehabilitation system as a means of communication to eliminate the spatial barriers between assistive device evaluation experts and remote disabled people. Although there are some references to how to use the telerehabilitation system itself as an assistive device evaluation and diagnosis system to promote the development of rehabilitation medicine, no substantial research has been conducted. Domestic products in this area include reports on the first national telerehabilitation system for the disabled developed by the Shenzhen Disabled Persons' Federation. The system focuses on the communication and exchange between experts and patients, allowing disabled people to consult experts on rehabilitation online and get rehabilitation advice.

　　Judging from the current development situation at home and abroad, the research of all parties has great limitations and is in the initial stage. Therefore, it is of great significance to conduct research on telerehabilitation systems.

　　In the telerehabilitation system, the information acquisition system is its main component. How to effectively control the information acquisition system at a distance, the quality of its implementation effect, and the speed of its implementation play a key role in the performance of the entire system. Since the telerehabilitation information acquisition system is a multivariable, nonlinear, time-varying system, it is difficult to establish an accurate mathematical model of the entire synchronous control system. Therefore, it is necessary to use an effective control method-fuzzy control.

2 Composition of telerehabilitation information acquisition control system

　　The schematic diagram of the telerehabilitation information acquisition control system is shown in Figure 1. The system is a robot for auxiliary camera, which can accept instructions to observe patients along a certain spatial curve as a path. This control system is mainly implemented by two functional modules. One is the PC at the on-site site, which receives the control command from the remote site through the Internet. After being processed by the fuzzy control algorithm, it is transmitted to the single-chip processing system through the RS 232 serial port to control the movement of the car, pan/tilt, and camera. In addition, the PC at the on-site site can also process the image information collected from the camera according to requirements, and then present it to the remote site in an appropriate manner through the Internet for diagnosis and design by remote rehabilitation experts and auxiliary design manufacturers. The second is the single-chip control system, which is mainly used to control the movement of the car, pan-tilt head, and camera so that they can reach the appropriate position, so that the remote rehabilitation experts can observe the patient's physical condition in real time without being restricted by time and space, and conduct remote diagnosis and evaluation. The single-chip control system can also process the signals of sensors such as the detection motor in place, and feedback the situation of the control fuzzy control system execution unit to the remote site. In simple terms, this fuzzy control system mainly realizes the automatic control of the car loaded with information acquisition devices, the pan-tilt head that drives the camera, and the movement of the camera, and collects real-time video or image information according to requirements for diagnosis and auxiliary product design.




3 Fuzzy control design of remote rehabilitation information acquisition system

3.1 Fuzzy control strategy of information acquisition system

　　The input variables of this system are: the steering angle from the car to the target, the distance from the car to the target, the height of the pan-tilt head from the target, the direction angle and distance between the camera and the target, a total of 6 input variables. The output variables are: the running speed and direction of the car's rudder motor, the running speed and direction of the car's drive motor, the running speed and direction of the motor that drives the pan-tilt head up and down, and the four directions of the pan-tilt head, a total of 10 output variables. Therefore, the initial control object of the information acquisition system has 6 input variables and 10 output variables, which belongs to the fuzzy controller of multi-input and multi-output structure.

　　Through fuzzy decoupling, this multi-input and multi-output fuzzy control structure is converted into a single variable fuzzy controller for design. The following takes the control of the speed of the trolley drive motor as an example to explain in detail the establishment of fuzzy control rules.

　　The trolley drive motor adopts a stepper motor, and its speed is controlled by changing the pulse frequency of the drive signal. Therefore, the control of the speed of the trolley drive motor adopts a single variable two-dimensional fuzzy controller, the input quantity is the error e of the distance from the trolley to the target and the rate of change of the error ec of the distance from the trolley to the target, and the output variable is the frequency f of the control pulse. In the specific implementation method of fuzzy control, the fuzzy table lookup method is adopted, and its principle is shown in Figure 2.




　　The error e and the error change rate ec obtained by each sampling are range-converted, that is, multiplied by the proportional factors k1 and k2, and then quantized, and the input physical signal value is converted into a point on the input domain, and the output control quantity can be obtained by querying the control action table. It is a point on the output domain, and then multiplied by the proportional factor k3 for range conversion, and the required control pulse frequency value f is obtained. The control action table is the correspondence between the points on the input domain and the output domain. It has gone through the process of fuzzification, fuzzy reasoning and defuzzification, and can be calculated offline. The table lookup method has a simple structure, is easy to implement, has low resource overhead, and runs online quickly. The

　　basic fuzzy subsets of error e, error change ec, and control quantity f are {NB (negative large deviation), NS (negative small deviation), 0 (zero), PS (positive small deviation), PB (positive large deviation)}. In the system, the domain of the error e from the car to the target distance is E, the domain of the error change rate ec from the car to the target distance is EC, and the domain of the output control quantity f is F. According to the actual situation of the system, its size is quantized into 5 levels, namely {-3, -1, 0, +1, +3), and the membership function curve shown in Figure 3 is selected. The controller can complete the fuzzification of the input variables. [page]




　　The fuzzy input variables are then inferred and decided by the fuzzy control rules to obtain the fuzzy output language variables {NB (negative large), NS (negative small), 0 (zero), PS (positive small), PB (positive large)}. Similarly, the output results inferred by the fuzzy controller must also be transformed into actual correction quantities to adjust the pulse frequency of the speed of the trolley drive motor to complete the control of the trolley speed.
　In order to simplify programming and facilitate real-time control, this system tabulates the control rules. The fuzzy controller is controlled according to the control state table shown in Table 1.






　　The selection of the quantization factors k1 and k2 of the error E and the error change rate EC has a great influence on the dynamic performance of the control system. k1 determines the response speed of the system. The larger k1 is, the faster the system responds, but the larger the overshoot is, and the longer the transition time is. k2 affects the overshoot of the system. The larger k2 is, the smaller the overshoot of the system will be, but the longer the response time of the system will be. k3 is the total gain of the fuzzy controller. If it is too small, the dynamic response process of the system will be prolonged, while if it is too large, the system will oscillate.

　　The control rules of other control quantities are similar to the control of the speed of the trolley drive motor mentioned above.

3.2 Software Design of Information Acquisition and Control System

　　At present, there are three technologies for constructing fuzzy controllers: using traditional single-chip microcomputers or microcomputers as the physical basis, compiling corresponding software to realize fuzzy reasoning and control; using single-chip microcomputers or integrated circuit chips to construct fuzzy controllers, and using configuration data to determine the structural form of fuzzy controllers; using programmable gate arrays to construct fuzzy controllers. Since the on-site site of the telerehabilitation system requires a microcomputer to receive remote control commands and process image information from cameras and transmit information through the Internet, in order to make full use of and save resources, we use microcomputers as the physical basis and compile corresponding software to realize fuzzy reasoning and control. The

　　upper computer software design of fuzzy control is mainly the design and implementation of fuzzy control algorithms, and also includes the design and implementation of the serial port communication part between the microcomputer and the single-chip microcomputer and the interface part with the Internet. The program flow is shown in Figure 4.




　　This part mainly realizes the fuzzy control function of the information acquisition system. Before the system runs, the host computer program must first be initialized and the serial port must be set to prepare for the correct operation of the system. When the remote control command is transmitted to the PC at the site through the Internet, it is processed by the fuzzy control algorithm and then sent to the single-chip control system through the serial port for execution. This control process does not require the personnel at the site to operate, and is completely remotely controlled. In this way, remote experts can easily control the operation of the information acquisition system according to their needs. At the same time, it is also convenient for local doctors or patients' families, reducing operational errors caused by communication barriers between remote experts and local doctors or families.

4 Conclusion

　　This system uses fuzzy control technology to solve the remote intelligent control of the remote rehabilitation information acquisition system, so that remote rehabilitation experts and auxiliary designers can conveniently remotely control the local information acquisition system through the Internet to accurately and real-time collect data information in a suitable way and angle for diagnosis and auxiliary product design. The experiment proves that the control system meets our design requirements and can remotely and real-timely collect three-dimensional visual information.