Control system of automatic defrosting robot for refrigeration house

Abstract

The invention discloses a control system of an automatic defrosting robot for a refrigeration house. The control system comprises a 3D visual detection module, a control module, a driving module and a defrosting robot body. The 3D visual detection module is used for collecting three-dimensional environment data in the refrigeration house in real time; the control module is used for receiving the three-dimensional environment data transmitted by the 3D visual detection module and fusing, analyzing and processing the three-dimensional environment data so as to determine the defrosting strategy and path of the defrosting robot body; the defrosting robot body comprises a walking mechanism and a defrosting executing mechanism, the driving module is used for driving motors of the walking mechanism and the defrosting executing mechanism of the defrosting robot body to operate, and the control system of the automatic defrosting robot for the refrigeration house can monitor the frosting condition in real time according to the refrigeration house environment; and the defrosting robot body can be started in time to conduct defrosting operation, the low efficiency of manual defrosting and the inaccuracy of timing defrosting are avoided, the defrosting efficiency is improved, the defrosting process is automatically completed by the control system, and the labor intensity of manpower is reduced.

Description

Technical Field

[0001] The present invention relates to the field of cold storage equipment automation, and more specifically, to a control system for an automatic defrosting robot in a cold storage. Background Art

[0002] During the operation of a cold storage, frost will continuously form on the surface of the evaporator. The presence of the frost layer will increase the thermal resistance, reduce the heat exchange efficiency of the evaporator, resulting in an increase in the energy consumption of the refrigeration system and a decline in the refrigeration effect. At present, the common defrosting methods include manual defrosting and timed defrosting. Manual defrosting has a high labor intensity, low efficiency, and it is difficult to defrost in a timely manner; timed defrosting cannot be adjusted according to the actual frosting situation, and problems such as untimely defrosting or over-defrosting may occur, resulting in energy waste.

[0003] Therefore, it is an urgent problem to be solved by the present invention to provide a control system for an automatic defrosting robot in a cold storage that can monitor the frosting situation of the evaporator in the cold storage in real time during use, control the automatic defrosting operation of the defrosting robot, improve the defrosting efficiency and accuracy, and reduce the operating cost of the cold storage. Summary of the Invention

[0004] In view of the above technical problems, the object of the present invention is to overcome the problems in the prior art that the common defrosting methods include manual defrosting and timed defrosting. Manual defrosting has a high labor intensity, low efficiency, and it is difficult to defrost in a timely manner; timed defrosting cannot be adjusted according to the actual frosting situation, and problems such as untimely defrosting or over-defrosting may occur, resulting in energy waste, so as to provide a control system for an automatic defrosting robot in a cold storage during use.

[0005] To achieve the above object, the present invention provides a control system for an automatic defrosting robot in a cold storage, the system includes: a 3D vision detection module, a control module, a driving module, and a defrosting robot body; wherein,

[0006] The 3D vision detection module includes: a radar and a vision sensor; it is used for real-time collection of three-dimensional environmental data in the cold storage;

[0007] The control module includes: a control center and a control algorithm unit; it is used for receiving the three-dimensional environmental data transmitted by the 3D vision detection module and performing fusion, analysis, and processing on the three-dimensional environmental data through the control algorithm unit to determine the defrosting strategy and path of the defrosting robot body;

[0008] The defrosting robot body includes: a traveling mechanism and a defrosting execution mechanism, the traveling mechanism enables the robot body to move in the cold storage and on the evaporator, and the defrosting execution mechanism is used for removing the frost layer on the surface of the evaporator;

[0009] The driving module is used to drive the motors of the traveling mechanism and the defrosting execution mechanism of the defrosting robot body.

[0010] Preferably, the radar is a lidar, and the vision sensor is a light field camera.

[0011] Preferably, the control algorithm unit fuses the three-dimensional environment data received from the 3D vision detection module through a radar-camera fusion method, and this method includes the following steps:

[0012] S1. Fix the millimeter-wave radar and the camera on the machine platform to ensure that the relative position between them remains unchanged. According to the camera imaging principle and the similarity principle, the relationship between P(u, v) and Q(X c , Y c , Z c ) is as follows:

[0013]

[0014] where P and Q are points on the image coordinate system and the camera coordinate system respectively, and the camera focal length f is 0 c and 0 is the distance on the Z c axis;

[0015] S2. Use the checkerboard and the Matlab camera calibration toolbox to calculate the camera internal parameters f x , f y , u0 and v0;

[0016] S3. Calibrate the rotation matrix R3×3 of the machine platform through a spirit level, and use OpenCV and the radar corner reflector to perform mean fitting in multiple sites to obtain the translation matrix T3×1 of the machine platform.

[0017] Preferably, the control algorithm unit further includes: a dual-input RPAM-YOLO target recognition algorithm to improve its running speed and detection accuracy; where

[0018] The network structure of the dual-input RPAM-YOLO target recognition algorithm includes: consisting of RPAM, backbone feature extraction, feature pyramid and decoupling module.

[0019] Preferably, the driving module includes: a navigation module, a motor driver and a drive controller; where

[0020] The navigation module is used to navigate the running trajectory of the defrosting robot body;

[0021] The driving controller is used to receive the motion state of the host computer, calculate the deflection angle of each walking motor and the rotation speed of the hub motor, and send the calculated data to each of the motor drivers. After receiving the data, each of the motor drivers executes the corresponding rotation of the walking motor to drive the walking mechanism of the defrosting robot body to move.

[0022] Preferably, the defrosting execution mechanism adopts a mechanical defrosting device, a hot gas defrosting device or an ultrasonic defrosting device.

[0023] Preferably, the walking mechanism adopts a crawler type or a wheel type structure and is equipped with an encoder.

[0024] Preferably, the system further includes: a wireless communication module; which is used to realize data transmission and instruction interaction between modules.

[0025] According to the above technical solution, the beneficial effects of the control system of the automatic defrosting robot for cold storage provided by the present invention when in use are as follows:

[0026] (1) The present invention uses a lidar and a vision sensor to realize the precise positioning of the defrosting robot in the cold storage. Among them, the lidar can obtain the environmental information around the defrosting robot body in real time and construct a map; the vision sensor can identify the characteristic markers in the cold storage to further improve the accuracy of positioning. Through the data fusion of these two sensors, the robot can accurately know its position and posture in the cold storage, so as to plan a reasonable defrosting path.

[0027] (2) The present invention receives and fuses the three-dimensional environment data transmitted by the 3D vision detection module through the control module and analyzes and processes the three-dimensional environment data through the control algorithm unit. When it is detected that the frosting on the evaporator surface reaches a certain degree, the control module will plan an optimal defrosting path according to the current position of the defrosting robot body and the position of the evaporator, and send a control instruction to the driving module of the defrosting robot body. After receiving the control instruction, the driving module is responsible for driving the movement and defrosting action of the defrosting robot body, so that the defrosting robot body moves to near or on the evaporator along the planned path, and controls the defrosting execution mechanism to perform defrosting operations on the evaporator surface. At the same time, the driving module can also adjust the movement speed and defrosting intensity of the robot according to the actual situation.

[0028] In summary, the present invention monitors the frosting situation in real time through the 3D vision detection module in the cold storage environment, and can start the defrosting robot body for defrosting operations in time, avoiding the inefficiency of manual defrosting and the inaccuracy of timed defrosting, and greatly improving the defrosting efficiency. And the entire defrosting process is automatically completed by the control system without manual intervention, reducing the labor intensity.

[0029] Other features and advantages of the present invention will be described in detail in the following specific implementation section; and parts not involved in the present invention are the same as or can be implemented using the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the following specific implementation, they are used to explain the present invention, but do not constitute a limitation to the present invention. In the drawings:

[0031] Figure 1 is a control principle block diagram of the control system of the cold storage automatic defrosting robot provided in a preferred embodiment of the present invention;

[0032] Figure 2 is a working principle block diagram of the traveling mechanism of the defrosting robot body of the control system of the cold storage automatic defrosting robot provided in a preferred embodiment of the present invention;

[0033] Figure 3 is a radar-camera coordinate schematic diagram of the control system of the cold storage automatic defrosting robot provided in a preferred embodiment of the present invention.

[0034] DESCRIPTION OF THE REFERENCE NUMERALS

[0035] 1, 3D vision detection module; 101, radar; 102, vision sensor; 2, control module; 201, control center; 202, control algorithm unit;; 3, drive module; 301, navigation module; 302, motor driver; 303, drive controller 4, defrosting robot body; 401, traveling mechanism; 402, defrosting execution mechanism; 5, wireless communication module. SPECIFIC IMPLEMENTATION

[0036] The following will describe in detail the specific implementation of the present invention with reference to the drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the present invention and does not limit the present invention.

[0037] As Figures 1-3 shown, a control system of a cold storage automatic defrosting robot provided by the present invention, the system includes: a 3D vision detection module 1, a control module 201, a drive module 3, and a defrosting robot body 4; wherein,

[0038] The 3D vision detection module 1 includes: a radar 101 and a vision sensor 102; and is used for real-time collecting three-dimensional environmental data in the cold storage;

[0039] The control module 2 includes: a control center 201 and a control algorithm unit 202; it is configured to receive the three-dimensional environment data transmitted by the 3D vision detection module 1 and perform fusion, analysis, and processing on the three-dimensional environment data through the control algorithm unit 202 to determine the defrosting strategy and path of the defrosting robot body 4.

[0040] The defrosting robot body 4 includes: a traveling mechanism 401 and a defrosting execution mechanism 402. The traveling mechanism 401 enables the robot body 4 to move inside the cold storage and on the evaporator, and the defrosting execution mechanism 402 is used to remove the frost layer on the surface of the evaporator.

[0041] The driving module 3 is configured to drive the motors of the traveling mechanism of the defrosting robot body 4 and the defrosting execution mechanism 5.

[0042] In the above solution, the 3D vision detection module 1 further includes: a frosting sensor, a temperature sensor, and a humidity sensor, which are used to monitor the frosting thickness, temperature, and humidity of the evaporator in the cold storage in real time; among them,

[0043] The frosting sensor is a capacitive sensor and is installed on the surface of the evaporator; the temperature sensor is installed on the defrosting robot body 4 or inside the cold storage; the temperature sensor uses an infrared sensor, and the humidity sensor uses a capacitive or resistive humidity sensor. The radar 101 and the light field camera 102 are installed on the defrosting robot body 4; the humidity sensor is installed inside the cold storage. The temperature of the surface of the evaporator and the humidity data of the surrounding air are collected in real time through the infrared sensor and the humidity sensor, and the data is transmitted to the control module 2. The control module 2 analyzes these data. When the changes in temperature and humidity reach the preset frosting threshold, it is determined that the frosting on the surface of the evaporator has reached the degree that requires defrosting.

[0044] In addition, the infrared sensor uses infrared thermal imaging technology for SLAM (Simultaneous Localization and Mapping). Infrared thermal imaging technology generates a temperature distribution image by capturing the infrared radiation emitted by the object itself and can maintain stable operation in extreme environments such as complete darkness, strong light glare, rain, fog, sand, and dust.

[0045] In the above solution, the radar 101 is a lidar; the vision sensor is a light field camera. In addition, the control module 2 selects an STM32F4 series chip. The STM32F4 series microcontroller not only includes single-cycle digital signal control center instructions, but also integrates a control center dedicated to floating-point operations. The STM32F407IGT6 memory has a flash memory of up to 1MB and provides 3 analog-to-digital converters, 2 digital-to-analog converters, 10 general-purpose timers, a low-power real-time clock, and a random number generator. The steering motor is a component that realizes the steering of the robot, with simple control and high steering accuracy. Currently, closed-loop stepper motors and servo motors are widely used in measuring the rotation angle. The rotation speed of the stepper motor depends on the pulse frequency of the command, while the rotation angle depends on the number of pulses given, and the output power is small, which can be used as a steering mechanism. The working modes of the encoder are divided into incremental and absolute.

[0046] During use, the lidar and the vision sensor are used to achieve precise positioning of the defrosting robot in the cold storage. Among them, the lidar can obtain the environmental information around the defrosting robot body 4 in real time and construct a map; the vision sensor 102 can identify the characteristic markers in the cold storage to further improve the accuracy of positioning. Through the data fusion of these two sensors, the robot can accurately know its position and posture in the cold storage, so as to plan a reasonable defrosting path.

[0047] The control module 2 receives and processes the three-dimensional environment data transmitted by the 3D vision detection module 1, and uses deep learning technology to fuse, analyze, and process the three-dimensional environment data through the control algorithm unit 202. When it is detected that the frosting on the evaporator surface reaches a certain degree, the control module 2 will plan an optimal defrosting path according to the current position of the defrosting robot body 4 and the position of the evaporator, and send a control instruction to the drive module 3 of the defrosting robot body. After receiving the control instruction, the drive module 3 is responsible for driving the movement and defrosting action of the defrosting robot body 4, so that the defrosting robot body 4 moves to near or on the evaporator along the planned path, and controls the defrosting actuator 402 to perform defrosting operations on the evaporator surface. At the same time, the drive module can also adjust the movement speed and defrosting intensity of the robot according to the actual situation.

[0048] In summary, the present invention monitors the frosting situation in real time through the 3D vision detection module 1 in the cold storage environment, and can timely start the defrosting robot body 4 to perform defrosting operations, avoiding the inefficiency of manual defrosting and the inaccuracy of timed defrosting, and greatly improving the defrosting efficiency. And the entire defrosting process is automatically completed by the control system without manual intervention, reducing the labor intensity.

[0049] In a preferred embodiment of the present invention, the control algorithm unit 202 fuses the three-dimensional environmental data received from the 3D vision detection module 1 through a radar-camera fusion method, and this method includes the following steps:

[0050] S1. Fix the millimeter-wave radar 101 and the camera on the machine platform to ensure that the relative position between the two remains unchanged. According to the camera imaging principle and the similarity principle, the relational expression between P(u, v) and Q(X c , Y c , Z c ) is as follows:

[0051]

[0052] where P and Q are respectively a point on the image coordinate system and the camera coordinate system, and the camera focal length f is the distance between 0 c and 0 on the Z c axis;

[0053] S2. Use the checkerboard and the Matlab camera calibration toolbox to calculate the camera internal parameters f x , f y , u0 and v0;

[0054] S3. Calibrate the rotation matrix R3×3 of the machine platform through the spirit level, and use OpenCV and the radar 101 corner reflector to perform mean fitting in multiple sites to obtain the translation matrix T3×1 of the machine platform.

[0055] In the above solution, by integrating the advantages of the radar and the camera, the performance of the system in target detection, positioning, and environmental perception is improved, and the safety and reliability are enhanced. The specific advantages include:

[0056] Information complementarity: The radar provides accurate distance and speed information, and the camera provides rich visual features. After fusion, the accuracy of target recognition and positioning is improved;

[0057] Accuracy improvement: By combining the data of both, the accuracy of target detection and tracking is improved, especially in complex environments;

[0058] Robustness enhancement: In bad weather or lighting conditions, the stability of the radar is combined with the high-resolution images of the camera to ensure the reliability of the system;

[0059] Cost reduction: Replace some expensive lidar functions with the radar to achieve a similar high-precision detection effect.

[0060] Through these advantages, the radar-camera fusion significantly improves the overall performance and application value of the system.

[0061] Therefore, when the defrosting robot body 4 moves in the real environment, it is also very important to identify the surrounding three-dimensional targets. By predicting the position, size, pose, and type of the three-dimensional targets, the defrosting robot body 4 can be assisted in adjusting its motion state and making corresponding path planning.

[0062] In a preferred embodiment of the present invention, the control algorithm unit 202 further includes: a dual-input RPAM-YOLO target recognition algorithm to improve its running speed and detection accuracy; wherein,

[0063] The network structure of the dual-input RPAM-YOLO target recognition algorithm includes: consisting of RPAM, backbone feature extraction, feature pyramid, and decoupling module.

[0064] In the above solution, the most advanced YOLO model currently is YOLOX proposed by Megvii Technology in 2021. Different from the previous generations of models based on the anchor-based strategy such as YOLOv3 and YOLOv4, YOLOX incorporates improvement strategies such as being anchor-free that have been intensively studied in the field of object detection in the past two years, making its running speed and detection accuracy exceed those of the previous generations of models. In order to better fuse radar information and camera information.

[0065] In addition, the main function of RPAM is to fuse the information collected by the radar and the visual images collected by the camera at the front end of the model, improving the efficiency of subsequent backbone feature extraction.

[0066] In a preferred embodiment of the present invention, the driving module 3 includes: a navigation module 301, a motor driver 302, and a driving controller 303; wherein,

[0067] The navigation module 301 is used to navigate the running trajectory of the defrosting robot body 4;

[0068] The driving controller 303 is used to receive the motion state of the upper computer, calculate the deflection angle of each walking motor and the rotation speed of the hub motor, and send the calculated data to each of the motor drivers 302. After receiving the data, each of the motor drivers 302 executes the corresponding rotation of the walking motor to drive the walking mechanism 401 of the defrosting robot body 4 to move.

[0069] In the above solution, the driving module 3 can be implemented by using pulse width modulation (PWM) technology. In addition, the driving module 3 is connected to the control module 2 by using a CAN bus or an Ethernet interface.

[0070] Among them, the walking motor further includes a steering motor, and the motor driver 302 further includes a steering motor driver.

[0071] After receiving a motion data signal from the CAN bus, each motor driver 302 controls the rotation of the motor through the motor control program for the data. And when each motor driver 302 receives the motion data signal, it can control the corresponding walking motor at the same time with high control precision.

[0072] In a preferred embodiment of the present invention, the defrosting actuator 402 adopts a mechanical defrosting device, a hot gas defrosting device or an ultrasonic defrosting device.

[0073] In the above solution, the mechanical defrosting device, the hot gas defrosting device or the ultrasonic defrosting device are all prior arts, and their structures and working principles will not be described in detail here.

[0074] In a preferred embodiment of the present invention, the traveling mechanism 401 adopts a crawler type or a wheel type structure and is equipped with an encoder.

[0075] In the above solution, the encoder selects a single-turn digital absolute encoder MAB25, which has strong seismic performance and is inexpensive.

[0076] In a preferred embodiment of the present invention, the system further includes: a wireless communication module 5; which is used to realize data transmission and instruction interaction between modules.

[0077] In the above solution, the wireless communication module 5 can adopt LoRa or ZigBee technology. The communication module 5 ensures smooth data transmission and communication between the control module 2 and the defrosting robot body 4. The control module 2 can monitor the running state, position information, defrosting effect, etc. of the defrosting control robot body 4 in real time.

[0078] In summary, the control system of the cold storage automatic defrosting robot provided by the present invention overcomes the common defrosting methods in the prior art, including manual defrosting and timed defrosting. Manual defrosting has a large labor intensity, low efficiency, and it is difficult to defrost in time; timed defrosting cannot be adjusted according to the actual frosting situation, and there may be problems such as untimely defrosting or over-defrosting, resulting in energy waste.

[0079] The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solutions of the present invention, and these simple modifications all belong to the protection scope of the present invention.

[0080] In addition, it should be noted that, among the various specific technical features described in the above specific embodiments, they can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not separately describe various possible combination manners.

[0081] In addition, any combination can be made among various different embodiments of the present invention, as long as it does not violate the idea of the present invention, and it should also be regarded as the content disclosed by the present invention.

Claims

1. A control system for an automatic defrosting robot in a cold storage, characterized in that, The system includes: a 3D vision detection module (1), a control module (201), a drive module (3), and a defrosting robot body (4); among them, The 3D vision detection module (1) includes: a radar (101) and a vision sensor (102); it is used to collect three-dimensional environmental data in the cold storage in real time; The control module (2) includes: a control center (201) and a control algorithm unit (202); it is used to receive the three-dimensional environmental data transmitted by the 3D vision detection module (1) and fuse, analyze, and process the three-dimensional environmental data through the control algorithm unit (202) to determine the defrosting strategy and path of the defrosting robot body (4); The defrosting robot body (4) includes: a traveling mechanism (401) and a defrosting execution mechanism (402). The traveling mechanism (401) enables the robot body (4) to move in the cold storage and on the evaporator. The defrosting execution mechanism (402) is used to remove the frost layer on the surface of the evaporator; The drive module is used to drive the motors of the traveling mechanism of the defrosting robot body (4) and the defrosting execution mechanism (5).

2. The control system of the automatic defrosting robot for cold storage according to claim 1, characterized in that, The radar (101) is a lidar, and the vision sensor (102) is a light field camera.

3. The control system of the automatic defrosting robot for cold storage according to claim 1 or 2, characterized in that, The control algorithm unit (202) fuses the three-dimensional environmental data received from the 3D vision detection module (1) through a radar-camera fusion method. This method includes the following steps: S1. Fix the millimeter-wave radar (101) and the camera on the machine table to ensure that the relative position between the two remains unchanged. According to the camera imaging principle and the similarity principle, the relational expression between P(u, v) and Q(X c , Y c , Z c ) is as follows: where P and Q are points on the image coordinate system and the camera coordinate system respectively, and the camera focal length f is 0 c and the distance from 0 on the Z c axis; S2. Calculate the internal parameters f, x f, y u0 and v0 of the camera by using the checkerboard and the Matlab camera calibration toolbox; x 、f y 、u0 and v0; S3. Calibrate the rotation matrix R3×3 of the machine platform through a level, and use OpenCV and the corner reflector of the radar (101) to perform mean fitting in multiple sites to obtain the translation matrix T3×1 of the machine platform.

4. The control system of the automatic defrosting robot for cold storage according to claim 1 or 3, characterized in that, The control algorithm unit (202) also includes: a dual-input RPAM-YOLO target recognition algorithm to improve its running speed and detection accuracy; among them, The network structure of the dual-input RPAM-YOLO target recognition algorithm includes: consisting of RPAM, backbone feature extraction, feature pyramid, and decoupling module.

5. The control system of the automatic defrosting robot for cold storage according to claim 1, characterized in that, The drive module (3) includes: a navigation module (301), a motor driver (302), and a drive controller (303); among them, The navigation module (301) is used to navigate the running trajectory of the defrosting robot body (4); The drive controller (303) is used to receive the motion state of the upper computer, calculate the deflection angle of each traveling motor and the rotation speed of the hub motor, and send the calculated data to each motor driver (302). After each motor driver (302) receives the data, it executes the corresponding rotation of the traveling motor to drive the traveling mechanism (401) of the defrosting robot body (4) to move.

6. The control system of the automatic defrosting robot for cold storage according to claim 1, characterized in that, The defrosting execution mechanism (402) adopts a mechanical defrosting device, a hot gas defrosting device, or an ultrasonic defrosting device.

7. The control system of the automatic defrosting robot for cold storage according to claim 1, characterized in that, The traveling mechanism (401) adopts a crawler or wheeled structure and is equipped with an encoder.

8. The control system of the automatic defrosting robot for cold storage according to claim 1, characterized in that, The system also includes: a wireless communication module (5); it is used to realize data transmission and instruction interaction between modules.

Images