Design of control system for vacuum cleaning robot

Introduction
With the aging of the population and the improvement of the social welfare system, the labor cost has risen sharply. Some simple repetitive physical labor has provided a broad market for service robots. Cleaning robots are a type of service robots that can replace people to clean rooms, workshops, walls, etc. A design scheme for a mobile cleaning robot used indoors is proposed. It has practical value. The main task of indoor cleaning robots is to replace people to clean, so they need to have a certain degree of intelligence. The cleaning robot should have the following capabilities: self-navigation, detection of walls, obstacles in the room and avoid them; ability to walk through most of the room, ability to detect the battery power and ability to return to charge autonomously, and at the same time require a relatively compact appearance, stable operation, and low noise; it should have a humanized interface for easy operation and control. Combined with the main functions of the cleaning robot, the hardware design of its control system is discussed.

1 Measurement and control system and functions
In order to make the vacuum cleaner robot move more smoothly and prevent it from getting stuck, the appearance of the vacuum cleaner robot is designed to be flat cylindrical. The flat circular design allows it to freely enter under sofas, beds and furniture, and clean some corners. The circular chassis parallel to the ground is supported by three wheels. The left and right sides are driving wheels, which are directly driven by two micro DC motors respectively, and the front support wheel is a universal wheel. This shape and wheel layout of the robot can easily realize turning on the spot, greatly improving the dexterity of walking, which is more prominent in places with a small space range. The multi-sensor system is composed of collision, ultrasonic and infrared sensors. An infrared receiving device is installed on the top of the robot; proximity sensors are installed every 45° on the bottom edge of the robot to detect steps and prevent falling; collision sensors are installed in front of the robot, and ultrasonic sensors are installed in front and on the left and right to detect the surrounding environment. The robot is equipped with a power management system. If the voltage is too low, it will stop cleaning and automatically charge.
1.1 Microcontroller
Traditional microprocessors such as the 51 series have a short development cycle and low cost, but their real-time performance is poor and it is difficult to implement complex control algorithms; in addition, the data conversion speed of the added peripheral circuit is slow, so the performance of the robot cannot be fully utilized. Although the emergence of high-speed DSP makes the system modular and fully digital, its development cost is high. ARM microprocessors with the same performance as DSP are rich in resources and have good versatility. Their main technical advantages are high performance, low price, and low power consumption. They are widely used in various fields. Therefore, it is a good strategy to apply ARM to robot control systems. LPC2210 is a Philips ARM7TDMI-S microprocessor with real-time simulation and tracking support. It uses a 3-level pipeline technology and can process instructions in parallel. Due to its very small size and extremely low power consumption, multiple 32-bit timers, PWM outputs and 32 GPIOs make it particularly suitable for industrial control and small robot systems, meeting the robot's requirements for controller operation speed. With LPC2210 as the core. Design a cleaning robot body system with simple structure and stable performance.
The robot control system mainly completes the following tasks: receiving data from sensors and encoders, and comprehensively processing and planning the cleaning path; driving the left and right wheels to move forward, controlling the cleaning and dust collection mechanisms, and completing various low-level control actions; designing a suitable human-machine interface, and displaying the robot status and running time on the LCD. Therefore, the robot control system includes a sensor module, a motor drive module, and a

Design of control system for vacuum cleaning robot

This ultrasonic sensor has the following advantages: adaptable to various environments, not affected by dust and light; blind area is 2.5 cm. The sensor can be installed in a suitable position to avoid the blind area; the detection divergence angle is 15°, the reaction distance is within 2.5 m, and the detection distance of this subject is 0.5 m. The basic principle of ultrasonic sensor is to measure the time taken from the emission of sound waves and the return to the receiver. The transmitting port and receiving port of this sensor are one pin. First, the controller emits a high-level pulse with a width of 5μs to stimulate the sensor to emit 40 kHz ultrasonic waves. After the pulse is emitted for 750μs, the pin level is set high; when the sensor receives the echo, the pin level is pulled low. The width of the high level at the signal end can tell the time required from emission to return, which is between 115μs and 18.5 ms. The formula s=vt/2, where s represents the distance between the sensor and the target; t represents the time from emission to recovery; v is the speed of sound waves, v=340 m/s. From this, the distance between the sensor and the obstacle can be known. The maximum detection time is 20 ms. It takes 100 ms for 5 sensors to query. Therefore, two adjacent sensors can be enabled in different time periods to avoid mutual interference and will not affect the speed of the robot.
1.2.2 Infrared proximity sensor The
reflective photoelectric switch is composed of infrared LED light source and photosensitive elements such as photodiode or photosensitive transistor. When there is an obstacle blocking, the light can be reflected back and the output is a low-level signal; when there is no obstacle blocking, the light cannot be reflected back and the output is a high-level signal. The
close-range infrared proximity sensor of the vacuum cleaner robot consists of two sets of identical infrared transmitting and receiving circuits. Each set of circuits can be divided into several parts, such as high-frequency pulse signal generation, infrared emission adjustment and control, infrared emission drive, and infrared reception. A 38 kHz modulated pulse signal is obtained through a 38 kHz crystal oscillator and a NOT gate circuit; the emission of the infrared transmitting tube (TSAL6200) is driven by a triode. The infrared light emitted by the transmitting tube is reflected by the object and received by the infrared receiving module. After being processed by the integrated circuit inside the receiving head (HS0038B), a digital signal is returned and input to the I/O port of the microcontroller, as shown in Figure 3. If the receiving head receives a 38 kHz infrared pulse, it will return a low level output, otherwise it will output a high level. By detecting the I/O port, the presence or absence of an object can be determined.
1.2.3 Collision switch sensor
Two slot-type photoelectric switches are evenly distributed on the left and right front of the robot. Such a layout allows the robot to sense obstacles from the front, left front, and right front directions, and thus respond differently according to the direction of the obstacle. When the robot encounters an obstacle, the spring presses the collision switch swing arm inward under the action of the obstacle, causing the reed to block the light of the photoelectric switch and output a low level. When there is no obstacle, the reed recovers under the action of the spring, the light of the photoelectric switch is not blocked, and the output is a high level, as shown in Figure 4.

Among these three sensors, the ultrasonic sensor is used to detect the walls and obstacles in front and on the left and right. The two ultrasonic sensors on the left and right are placed perpendicular to the walking direction and are used for the robot's walking planning along the edge; set the distance value between the robot and the wall when walking, adjust the robot's walking direction, so that the distance between the two ultrasonic waves and the wall is approximately equal to the set value, and keep the robot at an appropriate distance when walking along the wall, so that it will not collide with or move away from the wall. The two collision sensors in the front and one ultrasonic sensor are used to detect the environment in the front half; the contact sensor has the advantages of a large detection range, no signal conditioning, and less resource occupation. Through contact collision, it detects rod-shaped obstacles such as furniture legs that cannot be detected by the ultrasonic sensor. The positions between the sensors are shown in Figure 5.

The proximity sensor is used to detect whether there is a cliff on the ground. One is placed at the front, left front, right front and rear of the robot. In addition to the above three sensors, a normally open switch sensor is installed on each of the three wheels. When the wheel is suspended in the air, the switch will close and output a low level. When the wheel is suspended in the air, the robot can be stopped.

2 Motor control system
In low-power systems, DC motors have good linear characteristics and superior control performance, and are suitable for point and speed control. In order to achieve forward and reverse operation of the DC motor, it is only necessary to change the polarity of the motor power supply voltage. The change of voltage polarity and the length of the running time can be achieved by the processor, while the drive circuit is required to provide the current for the normal operation of the DC motor.
The H-bridge drive circuit is a more commonly used drive circuit. The two walking drive motors in this design are built with discrete device power field effect transistors and freewheeling diodes, which are low-cost and easy to dissipate heat, as shown in Figure 6.

Use ARM7's P0.8 and P0.9 to control the motor. These two pins are PWM output pins that can control the speed of the motor. This part mainly ensures that the robot can move in a plane. At the same time, the wheel is equipped with an encoder to detect the distance traveled. The robot can turn by dead reckoning. Assuming that the number of divisions of the robot's photoelectric encoder is N; the number of pulses received by the controller is m; the diameter of the wheel is D; the distance between the two wheels is W, then the distance the wheel moves forward is:

Assume that the robot's position in the environment coordinate system is (X(t), Y(t), φ(t)), then the azimuth φn+1 value of the n+1th sampling and the φ value of the nth sampling have the following relationship:

Where: vR(t) and vL(t) are the speeds of the two wheels at time t; △SLn and △SRn are the distances traveled by the two driving wheels from the nth sampling time to the n+1th sampling time.

If the requirement is to turn in place, that is, one wheel rotates forward and the other wheel rotates counterclockwise, then:

4 Conclusion
As a kind of service robot, cleaning robot has huge market potential and broad application prospects. With the development of sensor technology and the continuous advancement of microprocessors, prices are also falling. This study and design of a cleaning robot control system based on ARM7 microprocessor not only meets the requirements of practicality, but also provides good hardware support for software without increasing costs, and provides good technical support for better algorithms and software upgrades.