With the development of robot technology, automatic cleaning of curtain walls of high-rise buildings has become possible“ The “development of cleaning robot for complex arc curtain wall” is a project funded by the 863 program. The design task of the robot is to clean the metal and glass ceiling of the National Grand Theater located in Chang’an Street, Beijing. The theater is located in the center of Beijing and will become one of the landmark buildings in Beijing. In addition, due to the bad weather conditions in the north, the cleaning of the exposed walls of the theater is very important.
The robot system is applied to the cleaning operation of high-rise curtain wall, and the working conditions are poor. For engineering use, it undoubtedly requires the system to have higher safety and reliability. Therefore, the robot needs a good control system. According to the characteristics of high-altitude operation on curved surface, this paper introduces the mechanism composition of self climbing robot, and introduces the hardware system and software structure of robot control in detail.
2. Structural characteristics and robot design of National Grand Theatre
The main shape of the theater is semi ellipsoid, and the outer wall is covered by glass and titanium alloy plate, with a total surface area of 36000m2. The titanium plate on the Grand Theater is divided into 7 specifications, with a width ranging from 2.2m to 1.5m. Closed aluminum guide rails are distributed between each layer of titanium plates along the spherical weft direction, 25mm wide, 40mm higher than the plate surface, and the guide rails on glass and titanium plates are continuous. There are transverse and longitudinal gaps between titanium plates, in which a hemispherical structure for decoration is installed.
The scheme of self climbing robot is proposed based on the structural characteristics of buildings. According to the function, the robot body is divided into climbing mechanism, driving mechanism, cleaning mechanism and pitch adjustment mechanism. The robot prototype is shown in Figure 1, with a total length of about 3M, a width of 1m and a height of 0.5m. The body is made of light aluminum profiles. Through synchronous belt transmission, relative climbing motion is generated between the main frame and the rear box; The main drive is installed at the front end of the main frame, and the auxiliary drive is installed between two sliding guide rails with a length of 1000mm at the tail of the main frame. The auxiliary drive can slide passively within this range to adapt to the change of the length of the building slab; The main and auxiliary drives cooperate with the building guide rail when the robot moves, and provide driving force through friction wheels; The brush module can move vertically to the wall relative to the rear box and the main frame, and can move along the main frame synchronously with the rear box; The front and rear pitching support adjustment mechanism is mainly used to adjust the posture of the robot during climbing to adapt to the change of the angle between each floor of the building, and is used as a movable force fulcrum to improve the force when the robot climbs. The bottom of the main frame adopts a ship structure, which is matched with the shape of the sliding guide rod installed on the building guide rail, and further ensures the safety of robot climbing and movement under the coordination of the front and rear clamping parts. The structure of the front and rear clamping parts is the same. When the mechanism acts, the robot can firmly grasp the sliding guide rod and mechanically lock it to ensure the safety of the robot working at height.
The main technical indexes of the robot are as follows: the maximum working height is 50m; Maximum creeping speed 200mm / S; Cleaning efficiency ＞ 800m2 / day; The weight of the robot body is less than 150kg.
3. Design of control system for self climbing robot
Can bus is a serial data communication protocol developed by Bosch Company of Germany in the early 1980s to solve the data exchange between many control and test instruments in modern vehicles. Its excellent characteristics, high reliability and unique design are especially suitable for the interconnection of industrial process monitoring equipment. Therefore, it has attracted more and more attention in the industry, It has been recognized as one of the most promising fieldbus.
According to the motion function requirements and the robot mechanical body structure, the hardware system of the robot controller adopts the distributed can bus network structure, as shown in Figure 2. The system is divided into 6 parts and 5 can bus nodes: main frame common control node, rear box control node, climbing and pitching control node, main and auxiliary drive control node, main control computer node and remote control operation part.
3.1 control system structure distribution
As the central node, the main control computer (airborne IPC) becomes the core of the whole system. Pcm-9575 single board computer is adopted, which has powerful function, low power consumption and small volume. Embedded low-power via Ezra 800m processor can work up to 60 ℃ without fan, typical power consumption is 14W, and supports PC / 104 bus. Pcm3680 can interface adapter is selected to connect with pcm-9575 through PC / 104 bus, and adam-4550 transceiver module is used for wireless communication. The main control computer belongs to the high-level intelligent module and does not directly participate in the bottom control. It receives the motion instructions of the remote control operation box in real time, carries out motion planning and control scheduling for the robot, and sends instructions to the four bottom nodes such as the general controller of the main frame through the CAN bus to control and coordinate the work of the four nodes; On the other hand, the human body state information of the machine is fed back to the operator for monitoring. The main control computer is also equipped with an image processing card to process the video signal provided by the CCD camera on the robot to detect the cleanliness of the building guide rail and board surface.
The remote control operation box also uses adam-4550 module for wireless communication with the main control computer. The main function of the operation box is simple planning and monitoring. It receives and displays the image information collected by the wireless CCD camera at the same time. The operator monitors and intervenes the robot according to the image and the return information of the main control computer. The general controller of the main frame, the rear box controller, the climbing and pitching controller and the main and auxiliary drive controller belong to the bottom control level module, which is responsible for driving the corresponding DC motor. Each module is mainly coordinated through CAN bus communication. The four nodes and the man-machine interface circuit in the operation box are developed by ourselves. The distributed node controller is composed of P80C592 single chip microcomputer. The hardware structure of each controller is slightly different to realize the future expansion of the system function. P80C592 has rich input and output ports; 80c5l central processing unit (CPU) is adopted; External ROM can be extended to 64KB; 2 256 bytes on-chip RAM, external expandable to 64KB; The 10 bit ADC converter with 8-channel analog input and on-chip monitoring and tracking timer (WDT) fully meet the design requirements of node functions. At the same time, the built-in can controller increases the reliability of the system, which is more conducive to improve the integration and reduce the volume of the controller.
Take the climbing and pitching controller as an example, and the hardware structure principle block diagram of the controller node is shown in Figure 3. The climbing and pitching controller controls three DC servo motors of the climbing mechanism and the front and rear pitching support adjustment mechanism. The climbing controller encapsulates all the detection information necessary to complete the control function, mainly including the detection of the guide rail during climbing, the detection of the upper and lower limit positions of climbing movement, the detection of the contact state between the main frame and the sliding rod, the detection of the fit between the ship plate and the sliding rod notch, the detection of the parallelism between the main frame and the curtain wall, the detection of the front Contact detection and corresponding treatment of rear adjustment mechanism and curtain wall.
3.2 sensor configuration
The system uses a variety of sensors to detect the relative state between the robot and the environment and the movement of the body. Through the comprehensive processing and analysis (fusion) of sensor signals such as distance, material, obstacle and displacement, the controller realizes motion positioning, completes local autonomous intelligent control, and ensures effective scrubbing and smooth movement.
The configuration of robot external sensor is directly related to the working environment. The external information on the building mainly includes aluminum guide rail, sliding guide rod, decorative lamp, tall obstacle, large area dirt, a small amount of water stain, etc. According to the actual requirements of the operation, due to the climbing mechanism, in addition to the front and rear pitch adjustment support, the robot frame is in a non-contact state with the wall, and obstacles such as decorative lights do not affect the robot operation. A small amount of water stains will not cause adverse effects on scrubbing, and dirt and dust are the direct targets of scrubbing operation. Therefore, the external information processed by the system is mainly aluminum guide rail, sliding guide rod and tall obstacles. Through the detection of these three obstacles and the processing of the controller, the geometric reconstruction of the working environment is completed. In the control system, the contact switch and photoelectric sensor are used for the detection of tall obstacles, and the ultrasonic sensor and CCD camera are used for the detection of aluminum guide rail, sliding guide rod and suspension. In addition, the CCD camera enables the operator to understand the global state of the robot operation in real time through the method of wireless image transmission, so as to make necessary intervention.
The inner sensor is used to measure the state of the robot itself. The displacement sensor adopts incremental encoder, and uses the channel on the control board to input in interrupt mode; Proximity switches are used for the detection of various motion limit states. The energy part is an indispensable part of the system. The structure of the National Grand Theater is huge, and there are obstacles in the latitude direction of the building. For example, the towed cable will inevitably cause long cable length, heavy weight, easy damage, and may pollute and scratch the surface of the building. Therefore, lithium battery is selected as the robot energy, which has the advantages of cleanness, convenience, high energy storage efficiency and light weight. Therefore, the power monitoring part is also added to the detection system.
4. Robot control software
The working process of the robot is divided into four parts: self inspection, initialization, operation and information feedback. The software structure is shown in Figure 4.
The motion planning and operation management part makes a decision based on the comprehensive judgment of the operation environment information according to the requirements of the upper level planning and accurate local environment information, selects the most reasonable trajectory in response to the environment state for synthesis, and calls the motion module. The local environment model provided by the sensor detection information is the judgment basis for selection or scheduling. The motion control adopts modular design, and the motion modules are relatively independent. Each module can be coded, tested, debugged or modified independently, so as to simplify the complex work. All the movements required by the robot to complete the task can be combined by the basic action modules according to a certain logical relationship. After the planning, the software enters the output drive control stage, specifically distributes, executes and manages the action sequence, and finally forms various state information, which is used as the basis for fault diagnosis and processing. While the system processes, all operation results are output to the operation box and visualized.
This paper introduces the control system structure, controller, hardware composition and software structure characteristics of curtain wall cleaning robot in National Grand Theater. Engineering experiments show that it has high reliability, good performance and safe operation.
Responsible editor: GT