CN115373272A - Intelligent storage inspection robot control system and control method - Google Patents

Intelligent storage inspection robot control system and control method Download PDF

Info

Publication number
CN115373272A
CN115373272A CN202211076429.XA CN202211076429A CN115373272A CN 115373272 A CN115373272 A CN 115373272A CN 202211076429 A CN202211076429 A CN 202211076429A CN 115373272 A CN115373272 A CN 115373272A
Authority
CN
China
Prior art keywords
robot
inspection
module
platform
control
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211076429.XA
Other languages
Chinese (zh)
Inventor
邱秉钊
高升
张伟
李慧婕
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenyang University of Technology
Original Assignee
Shenyang University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shenyang University of Technology filed Critical Shenyang University of Technology
Priority to CN202211076429.XA priority Critical patent/CN115373272A/en
Publication of CN115373272A publication Critical patent/CN115373272A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
    • G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Manipulator (AREA)

Abstract

The invention provides an intelligent storage inspection robot control system and a control method, firstly, an embedded master control system is taken as a core, and the structure function of a robot body is perfected by combining a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module, a temperature and humidity module and the like; then, coordinating three application management layers of the intelligent inspection robot, wherein obstacles, smoke, flames and the like are identified in an inspection subsystem by adopting a method of directly predicting relative positions based on a YOLOv3 algorithm, so that the robot can carry out inspection obstacle avoidance and path planning more optimally; and finally, building various platforms of the inspection robot, and completing the track tracking of the inspection robot by adopting a self-adaptive robust sliding mode variable structure control algorithm in a system control submodule.

Description

Intelligent storage inspection robot control system and control method
Technical Field
The invention belongs to the technical field of intelligent storage, and particularly relates to a control system and a control method of an intelligent storage inspection robot.
Background
With the development of science and technology and the progress of society, the warehousing industry enters a high-speed development stage, the warehousing space area is also larger and larger, and the warehousing supervision technology is not synchronously developed, so that warehousing dangerous events in various places frequently occur, and huge loss is caused. Only depending on the manual management mode, no matter whether the storage personnel are sufficient or not, uninterrupted and error-free inspection within 24 hours is difficult to achieve, the human body is not sensitive to the perception of the environmental information, and the inspection personnel cannot easily find the environmental information when the environment is deteriorated. However, the following problems also exist with this type of management by means of device monitoring: monitoring devices installed at fixed points in a storage interval have monitoring blind areas; the mounted detection sensor is single, the pertinence is not strong, the missing report and false report rate is high, and the system is lack of intelligence; need multiple spot installation monitoring device, the cost is higher etc..
At present, the modes of manual management and equipment management are low in efficiency, high in cost, low in reliability and the like.
Disclosure of Invention
Therefore, the technical problem to be solved by the invention is to provide an intelligent storage inspection robot control system and a control method, which can solve the problems of low efficiency, high cost and poor reliability of the existing manual management and equipment management mode.
In order to solve the problems, the invention provides an intelligent storage inspection robot control system which comprises a robot body control structure unit, a robot application layer and a robot platform display;
the robot body control structure unit is used for receiving a control instruction of the robot application layer and realizing the direct shared control from the robot application layer to the robot body control structure unit;
the robot application layer is used for mapping and cooperating with each functional module of the robot body control structural unit and displaying the coordination plan of the human-computer interaction interface with the robot platform;
and the robot platform is displayed and used for monitoring the inspection process in real time by a worker in a man-machine interaction mode.
Optionally, the robot body control structure unit includes: the device comprises a detection signal conversion module, an embedded master control system, a body signal conversion module, a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module and a temperature and humidity module;
the embedded main control system is respectively in signal connection with the detection signal conversion module and the body signal conversion module, and the body signal conversion module is respectively in signal connection with the power supply module, the motor driving module, the alarm module, the ultrasonic module, the sensor module and the temperature and humidity module.
Optionally, the external signal passes through the detection signal conversion module and the body signal conversion module, and is transmitted to the embedded master control system after being converted;
the embedded main control system comprises a controller STM32, a power circuit, a crystal oscillator circuit, a reset circuit, an LED circuit and a JP1/2/3 wiring port;
the power supply circuit is connected with any one port of VDD _1/2/3 in the controller STM32, a USB interface in the power supply circuit is connected with 5V for power supply, and then the output voltage is 3.3V after voltage reduction processing;
the crystal oscillator circuit is connected with OSCIN and OSCOUT ports in the controller STM32 and used for providing clock signals for the chip;
the reset circuit is connected with an NRST port in the controller STM32, and when the NRST pin is pulled low, external reset is generated, and reset pulse is generated, so that the system is reset;
the LED circuit is connected with a VBAT port in the controller STM32 and used for supplying power to peripheral LEDs;
JP1/2/3 is used as wiring ports to be sequentially arranged, wherein JP1 is connected with SWIO and SWCLK ports in a controller STM 32; JP2/3 is connected with ports PA10 and PA9 in the controller STM 32.
Optionally, the robot application layer includes robot end management of the inspection subsystem, WEB end management of the network subsystem, and movable end management of the APP subsystem;
when the timing program triggers the robot to execute the inspection task and meets the internal inspection condition, the initialization program is started, various inspection hardware equipment is called, automatic navigation inspection is started according to task configuration requirements, and the flow and abnormal information can be checked in the APP subsystem by the monitoring personnel through data fusion of the inspection subsystem and the network subsystem and various hardware equipment.
Optionally, the robot platform display comprises an intelligent inspection platform, a big data platform, an internet of things platform and a mobile platform, and the intelligent inspection platform, the big data platform, the internet of things platform and the mobile platform are respectively responsible for inspection state monitoring, data storage management and human-computer interaction evaluation of the robot.
The invention provides a control method of an intelligent storage inspection robot, and the control system comprises a robot body control structure unit, a robot application layer and a robot platform display; the robot body control structure unit includes: the device comprises a detection signal conversion module, an embedded main control system, a body signal conversion module, a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module and a temperature and humidity module; the robot application layer comprises robot end management of the inspection subsystem, WEB end management of the network subsystem and movable end management of the APP subsystem; the robot platform display comprises an intelligent inspection platform, a big data platform, an Internet of things platform and a mobile platform;
the control method comprises the following steps:
the robot body is controlled to form a structural unit, and a robot application layer and a robot platform are bound with each other:
the method comprises the following steps: based on the robot body control structure unit, signal conversion of the inspection robot and coordination and consistent cooperation of all modules are completed, and a shared control instruction from a robot application layer is received;
step two: based on a robot application layer, performing layered management on an inspection robot subsystem, a network subsystem and an APP subsystem, and coordinating and controlling a body structure unit and transmitting information of a human-computer interaction interface;
step three: based on the robot platform display, an intelligent inspection platform, a big data platform, an internet of things platform and a mobile platform are built, and intelligent storage inspection robot control is achieved.
Optionally, the method for directly predicting the relative position based on the YOLOv3 algorithm identifies obstacles, smoke and flames, so that the robot performs inspection obstacle avoidance and path planning, and completes the trajectory tracking of the inspection robot by adopting a self-adaptive robust sliding mode variable structure control algorithm, and the algorithm realizes the steps of the storage inspection robot control method, thereby realizing the intelligent storage inspection of the robot.
Optionally, the method for directly predicting the relative position based on the YOLOv3 algorithm identifies the obstacle, smoke and flame, so that the robot performs inspection, obstacle avoidance and path planning, and the specific steps are as follows:
the method comprises the following steps: 2880 person data sets and 2474 picture data sets with smoke and flame, which are acquired by training and learning through an algorithm, wherein the two data sets comprise two scenes, namely indoor and outdoor;
step two: manually labeling the acquired data set by using labellimg software to generate an xml tag file based on the acquired data;
step three: the YOLOv3 network is utilized to train for 500epoch, and the effect of rapidly and accurately identifying obstacles, smoke and flames in the later period is achieved through multiple times of pre-training learning of the network structure.
Optionally, the trajectory tracking of the inspection robot is completed by adopting a self-adaptive robust sliding mode variable structure control algorithm, and the method specifically comprises the following steps:
the method comprises the following steps: establishing and providing a kinematic model of the inspection robot;
step two: designing an X position control law, a Y position control law and an attitude control law of the robot based on the mathematical description of the kinematics model, and proving the stability by using a Lyapunov method;
step three: the simulation effect analysis is further completed by using the results, and the feasibility and the robustness of the method are demonstrated.
Has the advantages that:
the embodiment of the invention provides a control system and a control method of an intelligent storage inspection robot, which comprises the following steps: the robot body controls a structural unit, an embedded main control system is used as a core of the robot, and the structural functions of the robot body are perfected by combining a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module, a temperature and humidity module and the like. The robot comprises a robot application layer, wherein the robot is mainly divided into three application layers: and the system comprises robot end management of the inspection subsystem, WEB end management of the network subsystem and movable end management of the APP subsystem. And the robot platform shows that the robot builds an intelligent inspection platform, a big data platform, an Internet of things platform and a mobile platform. The intelligent storage inspection robot control system is designed according to the scheme, and a method for directly predicting relative positions based on a YOLOv3 algorithm is adopted to identify obstacles, smoke, flames and the like, so that the robot can better perform inspection obstacle avoidance and path planning, and the track tracking of the inspection robot is completed by adopting a self-adaptive robust sliding mode variable structure control algorithm. The invention can realize the intelligent management of the warehousing system, thereby improving the efficiency, strengthening the automatic warehousing monitoring process, saving the manpower and material cost and ensuring the warehousing safety.
Drawings
Fig. 1 is a structural block diagram of an intelligent storage inspection robot control system according to an embodiment of the invention;
FIG. 2 is a schematic circuit diagram of a main control system of the intelligent storage inspection robot according to the embodiment of the invention;
FIG. 3 is a network structure diagram of the YOLOv3 algorithm according to the embodiment of the present invention;
FIG. 4 is a diagram of the actual effect of target detection based on the YOLOv3 algorithm according to the embodiment of the present invention;
FIG. 5 is a kinematic model diagram of an intelligent warehouse inspection robot according to an embodiment of the present invention;
fig. 6 is a diagram of a result of tracking a track of the intelligent warehouse inspection robot according to the embodiment of the invention.
Detailed Description
Referring to fig. 1 to 6 in combination, according to an embodiment of the invention, an intelligent warehouse inspection robot control system is provided. Firstly, an embedded main control system is taken as a core, and the structural functions of a robot body are perfected by combining a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module, a temperature and humidity module and the like; then, coordinating three application management layers of the intelligent inspection robot, wherein obstacles, smoke, flames and the like are identified in an inspection subsystem by adopting a method of directly predicting relative positions based on a YOLOv3 algorithm, so that the robot can carry out inspection obstacle avoidance and path planning more optimally; and finally, the inspection robot builds various platforms, and the track tracking of the inspection robot is completed in a system control submodule by adopting a self-adaptive robust sliding mode variable structure control algorithm. Compared with the defects of security measures of the warehousing system in the traditional mode, the control system can be widely applied to the warehousing field, and the intelligence of the inspection robot is realized through related algorithms, so that the control system has certain practical popularization value.
The invention provides an intelligent storage inspection robot control system and a control method, which replace part or all of manual inspection. The technical scheme is as follows: as shown in fig. 1;
in a first aspect: the inspection robot takes an embedded main control system as a core and is combined with a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module, a temperature and humidity module and the like to perfect the structural function of the robot body. The robot comprises a driving module, an alarm module, an ultrasonic module, a sensor module, a temperature and humidity module and the like, wherein the driving module, the alarm module, the ultrasonic module, the sensor module, the temperature and humidity module and the like are used for perfecting the structural function of a robot body; the body control structure unit decomposes each function module into the combination of motor action or other state signals to complete the specific execution function. The unit receives a control instruction of the robot application layer, and realizes direct shared control from the robot application layer to the body control structure unit.
Further, as shown in fig. 2, a part of the circuit connection diagram of the embedded master control system is shown, wherein the controller STM32, the power supply circuit, the crystal oscillator circuit, the reset circuit, the LED circuit and the JP1/2/3 wiring port are provided.
Further, the power supply circuit is connected with any one of VDD _1/2/3 ports in the controller STM32, namely the power supply circuit corresponds to the pin 24, the pin 36 and the pin 48 respectively, a USB interface in the power supply circuit is connected with 5V for power supply, and then the output voltage is 3.3V after voltage reduction;
the crystal oscillator circuit is connected with OSCIN and OSCOUT ports in the controller STM32, namely the crystal oscillator circuit corresponds to the pin 5 and the pin 6 respectively and is used for providing a clock signal for the chip;
the reset circuit is connected with an NRST port in the controller STM32, namely a pin 7, when the NRST pin is pulled low, external reset is generated, and a reset pulse is generated, so that the system is reset;
the LED circuit is connected with a VBAT port in a controller STM32, namely a pin 1, and is used for supplying power to peripheral LEDs;
JP1/2/3 are used as wiring ports which are sequentially arranged, wherein JP1 is connected with SWIO ports and SWCLK ports in the controller STM32, namely corresponding to the pin 34 and the pin 37 respectively; JP2/3 is connected with ports PA10 and PA9 in the controller STM32, namely corresponding to the pin 31 and the pin 30 respectively.
Further, the external detection signal and the robot body detection signal are converted and transmitted to a robot main control system so as to carry out the next action, and a robot main control chip adopts STM32 series in a microcontroller of ST (STMicroelectronics).
The robot body control structure unit includes: the device comprises a detection signal conversion module, an embedded master control system, a body signal conversion module, a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module and a temperature and humidity module;
the embedded main control system is in signal connection with the detection signal conversion module and the body signal conversion module respectively, and the body signal conversion module is in signal connection with the power module, the motor driving module, the alarm module, the ultrasonic module, the sensor module and the temperature and humidity module respectively.
Furthermore, the connection mode between each sub-module is mainly divided into two parts for coupling: and the robot body control unit is in signal connection and mechanical connection with the robot body structural unit.
Robot body control unit: firstly, a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module and a temperature and humidity module are coupled in a mechanical connection mode; secondly, the six modules not only carry out decision processing on the information from the body conversion module, but also transmit the information after decision processing to an embedded main control system of the robot structural unit in a signal connection mode. Robot body constitutional unit: the parameter information of the current environment is monitored in real time, after the effective data are obtained, the effective data are transmitted to the embedded main control system for processing through the detection signal conversion module, and the transmission mode between the effective data and the embedded main control system is signal connection; meanwhile, the embedded main control system sends the processed information to the body conversion module in a signal connection mode, and then the processed information flows to the robot body control unit.
Furthermore, the power supply circuit mainly comprises two parts, a power supply circuit (a common USB interface circuit can be used) and a voltage reduction circuit.
Furthermore, the key point of the design of the power part of the warehouse inspection robot system is a driving circuit module of the motor, and the motor driving module generally integrates module information in a mode of signal flow, such as a transistor module, a thyristor (silicon controlled) module, an IGBT module and the like. The system selects a drive chip TB6612FNG to drive a circuit to control a motor, so that the balance control, the speed control and the like of the robot are realized.
Furthermore, the alarm module is characterized in that the main control chip controls the buzzer module to sound and the sound frequency of the buzzer module through the analog signal after decision making. The alarm module mainly adopts a buzzer alarm circuit, and relates to temperature and humidity alarm, fire alarm, gas smoke concentration alarm and the like.
Furthermore, the ultrasonic module is triggered through the IO end of the main control chip, and the distance measurement is completed according to the time difference when the receiver receives the ultrasonic wave. The HC-SR04 ultrasonic ranging module can provide a non-contact distance sensing function, the ranging precision can reach 3mm, the module comprises an ultrasonic transmitter, a receiver and a control circuit, and the front obstacle can be found out in time in function, so that the inspection robot turns to in time and avoids the obstacle.
Further, the sensor module includes an infrared detection sensor, an attitude angle sensor: the two parallel functions can detect external environment information, observe the attitude angle of the robot body and finally transmit data through signals. The inspection robot can estimate the potential danger degree of the environment by detecting various parameters of the surrounding environment, and the most common detection method is to sense the temperature, humidity, toxic, flammable and explosive gas concentration in the environment by carrying a plurality of special-shaped sensors. Therefore, the multi-sensor fusion function is adopted to sense the environmental state change so as to facilitate the robot to make decisions.
An infrared detection sensor: the E18-D80NK is a photoelectric sensor integrating emission and reception, emitted light is emitted after modulation, and a receiving head demodulates and outputs reflected light, so that interference of visible light is effectively avoided. The sensor has the characteristics of long detection distance, small interference of visible light, low price, easiness in assembly, convenience in use and the like, and can be widely applied to various occasions such as robot obstacle avoidance, assembly line piece counting and the like.
Attitude angle sensor: an attitude solver is integrated in the module, and the current attitude of the module can be accurately output under a dynamic environment by matching with a dynamic Kalman filtering algorithm, the attitude measurement precision is static 0.05 degree, the dynamic 0.1 degree, the stability is extremely high, and the performance is even superior to that of some professional inclinometers.
Furthermore, the temperature and humidity module is generally designed and manufactured based on a unique humidity-sensitive capacitor unit and is packaged in a micro mode; the system has two signal communication modes of a single bus and a standard I2C. The temperature and humidity module adopts DHT12, has two communication modes of a single bus and standard I2C, has ultra-small volume and lower power consumption, and is suitable for various application occasions.
In a second aspect: the inspection robot is mainly divided into three application management layers: and the system comprises robot end management of the inspection subsystem, WEB end management of the network subsystem and movable end management of the APP subsystem. The barrier, smoke, flame and the like are identified by adopting a method for directly predicting the relative position based on the YOLOv3 algorithm, so that the robot can better perform routing inspection obstacle avoidance and path planning.
Furthermore, when the timing program triggers the robot to execute the inspection task and meets the internal inspection conditions, the initialization program is started, various inspection hardware devices are called, and automatic navigation inspection is started according to task configuration requirements. The layer mainly performs the following functions: first, the mapping and cooperation of each functional module of the subtask and body control structure unit mainly comprises gas detection, smoke and fire detection, temperature and humidity detection, voice alarm, path planning, autonomous obstacle avoidance, data fusion, autonomous return voyage and the like. And secondly, the coordination planning of the application management layer and the platform human-computer interaction interface mainly comprises various management of authority, tasks, batteries, data, robots, cameras, voice, navigation, video monitoring, abnormal events and the like.
Firstly, in the inspection process, different forms of detection are carried out on the storage environment, such as gas detection (oxygen content, hydrogen sulfide content, carbon monoxide content, methane content and the like), temperature and humidity detection, smoke/flame detection and the like, and if a certain index is different from the preset index, a voice alarm function is generated; in the inspection process, the robot plans a reasonable path, so that the inspection efficiency is improved; and if the collision foreign matter blocks the inspection process, automatically starting the autonomous obstacle avoidance function of the inspection robot.
Then, in the process, the inspection process can be managed through a WEB side, such as the problems of authority, tasks, battery charging and discharging, data storage, robot failure and the like.
And finally, through data fusion of the routing inspection subsystem and the network subsystem and various hardware devices (such as a high-definition camera, a voice module, a navigation module, a video display module and the like), the working personnel can check the flow and abnormal information in the APP subsystem.
In a third aspect: the inspection robot builds an intelligent inspection platform, a big data platform, an Internet of things platform and a mobile platform, and the track tracking of the inspection robot is completed by adopting a self-adaptive robust sliding mode variable structure control algorithm in a system control submodule. The robot platform is responsible for tasks such as inspection state monitoring, data storage management and human-computer interaction assessment of the robot, not only can indirectly control the robot body to control the structural unit, but also can assess certain interaction tasks through the robot application layer.
Based on the design idea of combining platform monitoring coordination control and local autonomous behavior control, the system structure of the control system is designed into a body control structure unit, an application layer and a platform display three-layer structure according to the hardware composition characteristics of the inspection robot system.
Furthermore, relevant software, hardware and parameter information of the robot are configured through the robot display platform, and a worker can monitor the inspection process in real time through the man-machine interaction mode.
An intelligent storage inspection robot is mainly divided into three application management layers. The inspection subsystem identifies obstacles, smoke, flames and the like by adopting a method for directly predicting relative positions based on a YOLOv3 algorithm, so that the robot can carry out inspection and obstacle avoidance more optimally.
The storage inspection robot has the visual detection capability which is an important embodiment of intellectualization, and the detection method of the visual image can comprehensively and deeply detect the storage environment, so that perfect information is provided for the decision of the inspection robot, and the task execution of the robot is efficient and accurate. The image recognition is not limited by the space size, can accurately position, quickly recognize and support various working environments, and increases the real-time performance and reliability of large-space detection. The YOLOv3 algorithm is adopted for target detection, and the network structure of the YOLOv3 algorithm is shown in fig. 3. YOLOv3 uses only convolution layers to make it a Full Convolution Network (FCN), proposes a feature extraction network Darknet-53: contains 53 convolutional layers, each followed by a batch normalization layer and a leak ReLU layer. And the pooling-layer-free layer replaces the pooling layer with the convolutional layer with the step of 2 to carry out the down-sampling process of the feature map, so that the loss of low-level features caused by the pooling layer can be effectively prevented.
Firstly, 2880 person data sets acquired by training and learning of the algorithm comprise an indoor scene and an outdoor scene, the data sets are manually labeled by labellimg software to generate an xml label file, and a Yolov3 network is used for 500epoch training. When the people walks around in the storage interval, when patrolling and examining the robot and meeting at certain within range, as long as patrolling and examining the robot within range, it can discern in real time and detect to send the control constitutional unit for the robot, accomplish each module of signal conversion back robot and can take certain pre-action, thereby supplementary and real-time let the robot make reasonable execution action or alarm, avoid colliding with and take place danger. Fig. 4 (a) shows the recognition and detection effect of the inspection robot on the scene.
Secondly, in the initial stage of fire in the storage space, when the temperature generally does not reach the ignition point, along with the generation of smoke, the YOLOv3 algorithm is used for identifying the smoke, alarming and processing actions are carried out in time, and the loss is reduced. 2474 pictures with smoke and flame indoors and outdoors are used at this time, collected pictures are labeled by using label, and 500epoch training is carried out on a YoloV3 network. The YOLOv3 algorithm is fast in real-time performance, and can rapidly and accurately identify smoke and further send out an alarm through early-stage massive pre-training learning so as to be processed in time, and the smoke identification effect is shown in fig. 4 (b). When the smoke reaches a certain threshold value, the inspection robot recognizes the smoke, immediately stops moving, immediately starts the application layer management system, starts the voice alarm device, and displays the smoke on the intelligent inspection platform through the human-computer interaction interface, so that the information of the storage section is transmitted.
In a word, the YOLOv3 algorithm experiment verifies that the inspection robot can effectively avoid pedestrians, detect and identify smoke and the like. The YOLOv3 algorithm intelligent detection and identification has unique advantages, the early warning capability of the warehouse inspection robot can be improved, the false alarm rate is reduced, and the accuracy of obstacle avoidance and fire warning is improved.
Furthermore, in a system control submodule of the intelligent storage inspection robot, the track tracking of the inspection robot is completed by adopting a self-adaptive robust sliding mode variable structure control algorithm, and the reliability of the storage inspection of the invention is verified.
First, a kinematic model of the inspection robot is given as shown in fig. 5: the two-wheeled differential patrol robot consists of two independent driving wheels, rotates around the same axis and is provided with one or more trundles, so that the robot keeps low friction force of a horizontal plane. The body control structure and the motor control are relatively simple, the robot has high flexibility, and the algorithm is easy to control. Therefore, the kinematics model of the warehouse inspection robot is described in the following formula (1):
Figure BDA0003831673760000081
wherein, [ x, y]In order to inspect the position of the robot,
Figure BDA0003831673760000082
is the derivative of x and is,
Figure BDA0003831673760000083
is the derivative of y; v is the linear velocity of the robot; the angle of orientation is represented by theta and,
Figure BDA0003831673760000084
is the derivative of θ; and w is the angular speed of the inspection robot.
Then, designing a self-adaptive robust sliding mode variable structure control algorithm:
for robot position [ x, y]Representing ideal locus by [ x ] d ,y d ]Describing, the derivative of the position error is known from (1) as the following equation (2):
Figure BDA0003831673760000091
wherein the content of the first and second substances,
Figure BDA0003831673760000092
is an ideal track [ x ] of the robot d ,y d ]A derivative;
Figure BDA0003831673760000093
for the robot position [ x, y]The error derivative of (2).
Further, designing an X position control law: let u 1 = vcos theta, and taking sliding mode variable structure function s 1 ＝x e To make a derivative
Figure BDA0003831673760000094
The X-axis control law is designed as follows:
Figure BDA0003831673760000095
is ready to obtain
Figure BDA0003831673760000096
Construction of Lyapunov function
Figure BDA0003831673760000097
And (5) obtaining a derivative:
Figure BDA0003831673760000098
then when k is 1 When the pressure is higher than 0, the pressure is higher,
Figure BDA0003831673760000099
indicating that the system is stable.
Wherein u is 1 Heel
Figure BDA00038316737600000910
The characterization meanings are the same; s 1 Representing the designed sliding mode function,
Figure BDA00038316737600000911
is s is 1 A derivative of (a); [ x ] of e ,y e ]Indicating robot position x, y]Ideal locus [ x ] of robot d ,y d ]The difference between them; k is a radical of formula 1 Is a constant greater than 0; v 1 Represents the designed Lyapunov function,
Figure BDA00038316737600000912
is a V 1 The derivative of (c).
Further, designing a Y position control law: let u 2 = vsin theta, taking sliding mode variable structure function s 2 ＝x e To make a derivative
Figure BDA00038316737600000913
The Y position control law is designed as follows:
Figure BDA00038316737600000914
is ready to obtain
Figure BDA00038316737600000915
Construction of Lyapunov function
Figure BDA00038316737600000916
Obtaining a derivative:
Figure BDA00038316737600000917
then when k is 2 When the pressure is higher than 0, the pressure is higher,
Figure BDA00038316737600000918
indicating that the system is stable.
Wherein u is 2 Heel
Figure BDA00038316737600000919
The characterization meanings are the same; s 2 Representing the designed sliding mode function,
Figure BDA00038316737600000920
is as s 2 A derivative of (a); [ x ] of e ,y e ]Indicating the robot position x, y]Ideal locus [ x ] of robot d ,y d ]The difference between them; k is a radical of 2 Is a constant greater than 0; v 2 Represents the designed Lyapunov function,
Figure BDA00038316737600000921
is a V 2 The derivative of (c).
Further, the actual position control law is obtained
Figure BDA00038316737600000922
In the formula: the actual control angle is described in θ. In order to make the control angle error be 0, an attitude control law needs to be designed.
Where v here denotes the actual position control law.
Further, designing an attitude control law: controlling the angular error theta e ＝θ-θ d Taking a sliding mode function s 3 ＝θ e Derivation of
Figure BDA00038316737600000923
The sliding mode attitude control law is as follows:
Figure BDA00038316737600000924
in the formula: xi 3 (t) = max (E (t)) + ξ represent the adaptive robust term, and ξ > 0 3 ＞0,E(t)＝0.5|s 3 |。
Construction of Lyapunov function
Figure BDA00038316737600000925
Derivation formula (3):
Figure BDA00038316737600000926
the solution is obtained from the Cauchy inequality:
Figure BDA0003831673760000101
then when k is 3 When the pressure is higher than 0,
Figure BDA0003831673760000102
indicating that the system is stable.
Wherein, theta e For controlling angle error, i.e. controlling angle theta and ideal angle theta d The difference value of (a) to (b),
Figure BDA0003831673760000103
is theta e A derivative of (a); s 3 For the designed sliding mode function,
Figure BDA0003831673760000104
is s is 3 A derivative of (a); k is a radical of formula 3 Is a constant greater than 0; v 3 Represents the designed Lyapunov function,
Figure BDA0003831673760000105
is a V 3 A derivative of (a); delta represents a sliding mode attitude control law; xi 3 (t) represents an adaptive robust term; ξ is a constant greater than 0; sgn is a mathematical sign function; | | represents modulo; e (t) denotes a sliding mode function s 3 The numerical relationship between them.
And finally, according to the implementation case, tracking the track of the storage inspection robot and judging the reliability of inspection. The analysis of the simulation results of the specific experiment is as follows:
in the storage space, the robot patrols and examines the direction and need be under system control, therefore the system can have a regulation orbit route, and the robot patrols and examines the platform and can show the robot in each direction information of x axle, y axle and patrol and examine the angle in real time.
1. Fig. 6 (a) can show that, in the initial stage of the experiment, there is a small error between the actual track and the ideal track of the inspection robot system, after the short-time operation, the actual track of the robot is basically fitted with the ideal track, no overlarge error occurs, the track tracking control effect is good, and the actual inspection effect of the inspection robot is good.
2. As can be seen from fig. 6 (b) and (c), the position tracking effect of the inspection robot [ x, y ] has a small error from the ideal position effect, and is basically fit. The inspection robot is proved to keep good stability in the pose, and the stability of the control system is reflected on the side surface.
3. As can be seen from fig. 6 (d), the robot is not at the initial position, and gradually approaches the target track after a while under the control strategy. The deviation of the course angle has small fluctuation, and the angle output of the robot is stable. And the stability can be kept better under the condition that unknown disturbance exists and speed jitter is generated.
The problem of tracking the track of a patrol robot is solved: firstly, a kinematic model of the robot is established, an error model is established by means of the kinematic model, and an error equation is deduced according to the error model; then, considering the situations of low convergence speed, more jitter and the like of a classical controller, a new position and attitude control law is provided; finally, the stability of the designed control law was demonstrated with the Lyapunov method. Simulation results show that the designed inspection robot can effectively track a given ideal track, errors can be effectively reduced, and the robustness of the system is kept.
The embodiment of the invention can be realized by a software algorithm and a necessary universal hardware platform. The robot body control structure unit, the robot application layer and the robot platform are mutually connected to form a control system, and obstacles, smoke, flames and the like are identified by adopting a method for directly predicting relative positions based on a YOLOv3 algorithm, so that the robot can better perform routing inspection and obstacle avoidance and path planning and complete track tracking of the routing inspection robot by adopting a self-adaptive robust sliding mode variable structure control algorithm, and intelligent management of a storage system can be realized, so that the efficiency is improved, the automatic storage monitoring process is strengthened, the labor and material costs are saved, and the storage safety is ensured.
It is readily understood by a person skilled in the art that the advantageous ways described above can be freely combined, superimposed without conflict.

Claims (9)

1. An intelligent storage inspection robot control system is characterized by comprising a robot body control structure unit, a robot application layer and a robot platform display;
the robot body control structure unit is used for receiving a control instruction of the robot application layer and realizing the direct shared control from the robot application layer to the robot body control structure unit;
the robot application layer is used for mapping and cooperating with each functional module of the robot body control structural unit and displaying the coordination plan of the human-computer interaction interface with the robot platform;
and the robot platform is displayed and used for monitoring the inspection process in real time by a worker in a man-machine interaction mode.
2. The intelligent warehouse inspection robot control system according to claim 1, wherein the robot body control structure unit comprises: the device comprises a detection signal conversion module, an embedded main control system, a body signal conversion module, a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module and a temperature and humidity module;
the embedded main control system is respectively in signal connection with the detection signal conversion module and the body signal conversion module, and the body signal conversion module is respectively in signal connection with the power supply module, the motor driving module, the alarm module, the ultrasonic module, the sensor module and the temperature and humidity module.
3. The intelligent storage inspection robot control system according to claim 2, wherein external signals pass through the detection signal conversion module and the body signal conversion module, are converted and then are transmitted to the embedded master control system;
the embedded main control system comprises a controller STM32, a power circuit, a crystal oscillator circuit, a reset circuit, an LED circuit and a JP1/2/3 wiring port;
the power supply circuit is connected with any one port of VDD _1/2/3 in the controller STM32, a USB interface in the power supply circuit is connected with 5V for power supply, and then the output voltage is 3.3V after voltage reduction processing;
the crystal oscillator circuit is connected with OSCIN and OSCOUT ports in the controller STM32 and used for providing a clock signal for the chip;
the reset circuit is connected with an NRST port in the controller STM32, and when the NRST pin is pulled low, external reset is generated, and reset pulse is generated, so that the system is reset;
the LED circuit is connected with a VBAT port in the controller STM32 and used for supplying power to peripheral LEDs;
JP1/2/3 are used as wiring ports to be sequentially arranged, wherein JP1 is connected with SWIO and SWCLK ports in the controller STM 32; JP2/3 is connected with ports PA10 and PA9 in the controller STM 32.
4. The intelligent warehouse inspection robot control system according to claim 1, wherein the robot application layer comprises robot end management of the inspection subsystem, WEB end management of the network subsystem and movable end management of the APP subsystem;
when the timing program triggers the robot to execute the inspection task and meets the internal inspection condition, the initialization program is started, various inspection hardware devices are called, automatic navigation inspection is started according to task configuration requirements, and the monitoring personnel can check the flow and abnormal information in the APP subsystem through data fusion of the inspection subsystem and the network subsystem and the various hardware devices.
5. The intelligent storage inspection robot control system according to claim 1, wherein the robot platform display comprises an intelligent inspection platform, a big data platform, an internet of things platform and a mobile platform, and the intelligent inspection platform, the internet of things platform and the mobile platform are respectively responsible for inspection state monitoring, data storage management and human-computer interaction evaluation of the robot.
6. An intelligent storage inspection robot control method is characterized in that the intelligent storage inspection robot control system according to any one of claims 1 to 5 comprises a robot body control structure unit, a robot application layer and a robot platform display; the robot body control structure unit includes: the device comprises a detection signal conversion module, an embedded main control system, a body signal conversion module, a power supply module, a motor driving module, an alarm module, an ultrasonic module, a sensor module and a temperature and humidity module; the robot application layer comprises robot end management of the inspection subsystem, WEB end management of the network subsystem and movable end management of the APP subsystem; the robot platform display comprises an intelligent patrol platform, a big data platform, an Internet of things platform and a mobile platform;
the control method comprises the following steps:
the robot body controls the structural unit, the robot application layer and the robot platform to be bound with each other:
the method comprises the following steps: based on the robot body control structure unit, signal conversion of the inspection robot and coordination and consistent cooperation of all modules are completed, and a sharing control instruction from a robot application layer is received;
step two: based on a robot application layer, performing layered management on an inspection robot subsystem, a network subsystem and an APP subsystem, and coordinating and controlling a body structure unit and transmitting information of a human-computer interaction interface;
step three: based on the robot platform display, an intelligent inspection platform, a big data platform, an internet of things platform and a mobile platform are built, and intelligent storage inspection robot control is achieved.
7. The intelligent storage inspection robot control method according to claim 6, wherein the method for directly predicting the relative position based on the YOLOv3 algorithm identifies obstacles, smoke and flames, so that the robot performs inspection obstacle avoidance and path planning and track tracking of the inspection robot is completed by adopting a self-adaptive robust sliding mode variable structure control algorithm, and the algorithm realizes the steps of the storage inspection robot control method, thereby realizing the intelligent storage inspection of the robot.
8. The intelligent storage inspection robot control method according to claim 7, wherein the method for directly predicting the relative position based on the YOLOv3 algorithm identifies obstacles, smoke and flames, so that the robot performs inspection obstacle avoidance and path planning, and the specific steps are as follows:
the method comprises the following steps: 2880 person data sets and 2474 picture data sets with smoke and flame, which are acquired by training and learning through an algorithm, wherein the two data sets comprise two scenes, namely indoor and outdoor;
step two: manually labeling the acquired data set by using labelimg software to generate an xml tag file based on the acquired data;
step three: the YOLOv3 network is utilized to train for 500epoch, and the effect of rapidly and accurately identifying obstacles, smoke and flames in the later period is achieved through multiple times of pre-training learning of the network structure.
9. The intelligent storage inspection robot control method according to claim 7, wherein the trajectory tracking of the inspection robot is completed by adopting a self-adaptive robust sliding mode variable structure control algorithm, and the method comprises the following specific steps:
the method comprises the following steps: establishing and providing a kinematics model of the inspection robot;
step two: designing an X position control law, a Y position control law and an attitude control law of the robot based on the mathematical description of the kinematics model, and proving the stability by using a Lyapunov method;
step three: the simulation effect analysis is further completed by using the results, and the feasibility and the robustness of the method are demonstrated.
CN202211076429.XA 2022-09-05 2022-09-05 Intelligent storage inspection robot control system and control method Pending CN115373272A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211076429.XA CN115373272A (en) 2022-09-05 2022-09-05 Intelligent storage inspection robot control system and control method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211076429.XA CN115373272A (en) 2022-09-05 2022-09-05 Intelligent storage inspection robot control system and control method

Publications (1)

Publication Number Publication Date
CN115373272A true CN115373272A (en) 2022-11-22

Family

ID=84069821

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211076429.XA Pending CN115373272A (en) 2022-09-05 2022-09-05 Intelligent storage inspection robot control system and control method

Country Status (1)

Country Link
CN (1) CN115373272A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117686670A (en) * 2024-02-02 2024-03-12 内蒙古蒙牛乳业（集团）股份有限公司 Sample automatic detection system and intelligent laboratory

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117686670A (en) * 2024-02-02 2024-03-12 内蒙古蒙牛乳业（集团）股份有限公司 Sample automatic detection system and intelligent laboratory
CN117686670B (en) * 2024-02-02 2024-05-03 内蒙古蒙牛乳业（集团）股份有限公司 Sample automatic detection system and intelligent laboratory

Similar Documents

Publication Publication Date Title
WO2021196529A1 (en) Air-ground cooperative intelligent inspection robot and inspection method
CN102096413B (en) Security patrol robot system and control method thereof
CN110673603B (en) Fire scene autonomous navigation reconnaissance robot
CN106864739A (en) A kind of six rotor flying robots for underground pipe gallery detection
CN108297058A (en) Intelligent security guard robot and its automatic detecting method
CN112454353B (en) Inspection robot and inspection method for detecting leakage of dangerous gas
CN113325837A (en) Control system and method for multi-information fusion acquisition robot
CN113189977B (en) Intelligent navigation path planning system and method for robot
CN111624641A (en) Explosion-proof type intelligent inspection robot for oil depot area
CN105807760A (en) Intelligent robot, method of intelligent robot of automatically planning paths, and device
CN110103224A (en) A kind of harbour security protection crusing robot and control system
CN104953709A (en) Intelligent patrol robot of transformer substation
CN113730860A (en) Autonomous fire extinguishing method of fire-fighting robot in unknown environment
CN115373272A (en) Intelligent storage inspection robot control system and control method
Merriaux et al. The vikings autonomous inspection robot: Competing in the argos challenge
CN115246121A (en) Patrol robot and patrol robot system
CN109857121A (en) Indoor inspection mobile robot
CN115599103A (en) Explosion-proof intelligent inspection robot system for petrochemical field
CN112286190A (en) Security patrol early warning method and system
CN215281955U (en) Building security patrol robot
CN113625773A (en) Unmanned aerial vehicle emergency fire-fighting forest patrol command system
CN113246152A (en) Intelligent inspection robot for underground substation for mine
CN212963694U (en) Patrol robot
CN105892454A (en) Mobile intelligent security check robot
CN214474623U (en) 5G tobacco house environmental monitoring intelligent vehicle based on linear CCD tracking

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination