CN102298375A - Real-time humanoid robot control system and method - Google Patents

Real-time humanoid robot control system and method Download PDF

Info

Publication number
CN102298375A
CN102298375A CN2011101855130A CN201110185513A CN102298375A CN 102298375 A CN102298375 A CN 102298375A CN 2011101855130 A CN2011101855130 A CN 2011101855130A CN 201110185513 A CN201110185513 A CN 201110185513A CN 102298375 A CN102298375 A CN 102298375A
Authority
CN
China
Prior art keywords
network
real
controller
time
upper motion
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
CN2011101855130A
Other languages
Chinese (zh)
Inventor
孙逸超
熊蓉
冯冬芹
褚健
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.)
Zhejiang University ZJU
Original Assignee
Zhejiang University ZJU
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 Zhejiang University ZJU filed Critical Zhejiang University ZJU
Priority to CN2011101855130A priority Critical patent/CN102298375A/en
Publication of CN102298375A publication Critical patent/CN102298375A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • YGENERAL 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
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02PCLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Landscapes

  • Manipulator (AREA)
  • Numerical Control (AREA)

Abstract

The invention relates to a real-time humanoid robot control system and a method. The real-time humanoid robot control system comprises an upper motion controller, a plurality of network controllers and a plurality of end nodes. A dual-layer network structure is adopted. The upper motion controller and the network controllers form an upper-layer network which carries out communication through a real-time Ethernet network. Each network controller and the end nodes form a lower-layer network which carries out communication through a communication method adopted by the end nodes. By using the real-time Ethernet network to connect the upper motion controller with the network controller, on the premise that the real-time performance is ensured, the bandwidth of the system is widened and general end nodes which adopt a non-real-time Ethernet network for communication are compatible. By widening the bandwidth of the system, the network signal transmission speed and the transmission capability of the system are improved, the entire system can transmit more control commands at higher speed, and the control accuracy and the control smoothness are improved.

Description

A kind of anthropomorphic robot real-time control system and method
Technical field
The invention belongs to the robot field, relate to a kind of anthropomorphic robot control system and method, exactly, relate to a kind of anthropomorphic robot Control Network structure and control method based on real-time ethernet.
Technical background
Anthropomorphic robot is a kind of robot that imitates the skeleton structure, and it and other machines people's maximum difference is the foot sport mode of its apery.Show that from bionical angle analysis anthropomorphic robot has advantages such as accommodative ability of environment is strong, flexible movements, therefore, has more wide application and development prospect.
Anthropomorphic robot will be realized the kinematic dexterity and the motion smoothing of height, and needs real-time nearly 30 motors of control to realize multiple motion.This all needs robot to have the Control Network of strong, reliable, real-time and high bandwidth, makes accurately and motion response timely with environment to external world.
Present existing anthropomorphic robot mainly is divided into two classes on control mode, centralized control mode and distributed control mode.The early stage anthropomorphic robot P2 of Japan HONDA company adopts the Centralized Control System of VME (Versa Module Eurocard) bus, and Yan Fa anthropomorphic robot ASIMO then adopted the dcs based on PCI-E (PCI Express) bus design afterwards.The SDR-3X/4X of Japan Sony Corporation then is based on open universal serial bus OPEN-R technological development.
Anthropomorphic robot generally requires to carry polytype sensor acquisition environment and oneself state information as the input of robot control algorithm and the foundation of assessment robotary because environmental information is required complexity.Various kinds of sensors is on the market often used different communication interfaces at present, as CAN bus, 485 buses, serial line interface etc.Different hold facilities different interfaces, cause the hardware system complexity, be difficult for carrying out unified debugging and control.
Real-time ethernet refers to be based upon on the basis of IEEE802.3 standard, improves real-time by the real-time extension to itself and relevant criterion, and accomplishes the Ethernet standard with the standard ethernet seamless link.In May, 2003, IEC/SC65C has set up WG11 real-time ethernet working group specially, is responsible to define IEC 61784-2 " based on industrial communication network professional etiquette in the real-time application system of ISO/IEC 8802-3 " international standard.This standard comprises 11 kinds of real-time ethernet professional etiquette collection such as Communication Profile Family 2 Ethernet/IP, CPF3 PROFINET, CPF4 P-NET, CPF6 Interbus, CPF10 VNET/IP, CPF11 TCNET, CPF12 EtherCAT, CPF13 Ethernet Powerlink, CPF14 EPA, CPF15 Modbus/TCP and CPF16 SERCOS.
" the EPA(Ethernet for Plant Automation that is used for commercial measurement and control system) system architecture and communication standard " be a kind of based on information network communication technologys such as Ethernet, WLAN (wireless local area network), bluetooths, be applicable to the industrial control network communication standard of industrial automation control system device and instrument and meter intercommunication.The EPA standard is a kind of Industrial Ethernet real-time communication system standard that China proposes, adopt advanced switching technology, vlan technology, processed technology, and formulate real-time communication protocol to improve the Ethernet real-time performance in application layer, neither change the Ethernet original structure, kept simple, the inexpensive advantage of Ethernet, the requirement of the real-time that reaches simultaneously becomes the present main means of improving the Industrial Ethernet real-time performance.
Chinese patent publication number CN101293350A has invented a kind of apery robot distributed dual-bus motion control system, employing is based on the distributed system of CAN bus, and system is made of coordinated movement of various economic factors controller, CAN bus, local bus and some joint controls.Tuning controller links to each other with all joint controls by the CAN bus, and is interconnected by the local bus realization between all or part of joint control, shares joint status information and sensor information.But along with the increase and the raising of exercise data update cycle of joint of robot quantity, the bandwidth of the highest only 1mbps of CAN bus between motion controller and joint control can't satisfy the demand of data traffic.
Summary of the invention
The objective of the invention is at the deficiency of prior art in the existence of anthropomorphic robot application scenario, a kind of anthropomorphic robot real-time control system and method are provided, on the basis of satisfying anthropomorphic robot real-time and reliability requirement, provide higher bandwidth and transmission speed, simultaneously compatible existing various end-equipments.
The objective of the invention is to be achieved through the following technical solutions: a kind of anthropomorphic robot real-time control system, it comprises a upper motion controller, plurality of network controller and several endpoint nodes; Adopt the double-layer network structure, the plurality of network controller links to each other with upper motion controller by real-time ethernet respectively, forms a upper layer network and carries out communication; Each network controller links to each other with several endpoint nodes respectively; Form lower floor's network and carry out communication.
Further, described real-time ethernet network is Industrial Ethernet EPA.
Further, described network controller links to each other with several endpoint nodes respectively by CAN bus, 485 buses or serial line interface.
Further, described endpoint node is motor driver, pot, plantar pressure sensor, power and torque sensor or attitude sensor.
A kind of anthropomorphic robot real-time control method of using said system, this method comprise following two processes:
(1) upper motion controller need be when a certain endpoint node sends instruction or data, upper motion controller at first will instruct or data send to the network controller of this endpoint node correspondence by upper layer network, network controller is resolved the packet that receives, extract data division and encapsulate, by lower floor's network the packet that encapsulation obtains is sent to the target endpoint node respectively then according to the form of lower floor's network;
(2) endpoint node need be when upper motion controller send state information and feedback information, at first packet is sent to coupled network controller by lower floor's network, network controller resolution data bag content, extract data division and it is encapsulated as the data layout of real-time ethernet, the packet that encapsulation is obtained by upper layer network sends to upper motion controller.
The invention has the beneficial effects as follows, by using real-time ethernet to connect upper motion controller and network controller, guaranteeing to have enlarged system bandwidth on the basis of real-time and reliability, the general endpoint node that compatible again employing non real-time Ethernet carries out communication.By enlarging system bandwidth, improved grid signaling rate and transmittability, make total system can transmit the more control instruction at faster speed, improved the precision and the flatness of control.
Description of drawings
Fig. 1 is an anthropomorphic robot real-time control system structural representation;
Fig. 2 is the structural representation of anthropomorphic robot real-time control system example 1;
Fig. 3 is a data downstream step synoptic diagram in the anthropomorphic robot real-time control method;
Fig. 4 is a data uplink step synoptic diagram in the anthropomorphic robot real-time control method.
Embodiment
Anthropomorphic robot real-time control system of the present invention comprises a upper motion controller, plurality of network controller and several endpoint nodes.Adopt the double-layer network structure, the plurality of network controller links to each other with upper motion controller by real-time ethernet respectively, forms a upper layer network, carries out communication.Each network controller links to each other with several endpoint nodes respectively; Form a plurality of lower floors network, adopt the communication modes that endpoint node adopted to carry out communication.
Upper motion controller is responsible for providing control signal and data, sends to network controller, and receives the data message that the network controller transmission comes.Upper motion controller can be realized by PC, DSP or the embedded system based on the x86 framework, but is not limited thereto.
Network controller is responsible for receiving control command and the data from upper motion controller, be forwarded to each endpoint node, the data message that will be received from endpoint node simultaneously sends to upper motion controller, network controller can compatible existing all kinds of endpoint node equipment communication interface, as CAN bus, 485 buses, serial line interface etc., still the communication interface with upper motion controller is unified; Network controller can adopt the embedded system of ARM system, but is not limited thereto.
Endpoint node is responsible for receiving and carry out the control command that network controller sends, and status information and feedback information are sent to network controller.Endpoint node can be motor driver, pot, plantar pressure sensor, power and torque sensor or attitude sensor.
Connected mode between described network controller and the endpoint node can be CAN bus, 485 buses or serial line interface.
Described real-time ethernet network is Industrial Ethernet EPA.EPA is a kind of based on information network communication technologys such as Ethernet, WLAN (wireless local area network), bluetooths, is applicable to the industrial control network communication standard of industrial automation control system device and instrument and meter intercommunication.EPA has guaranteed the requirement of Industry Control to real-time and reliability in transmission bandwidth that has Ethernet and speed.All kinds of ripe bus system that is widely used in robot and motion control field at present comprises CAN bus, RS485 bus etc.With the CAN bus is example, and its maximum transmission bandwidth is 1Mbps.In Motor Control Field, the typical CAN data packet length that comprises a position command is at least 20 bytes, if will be that control cycle carries out position control with 4ms to 20 motors, need to consume the bus bandwidth of 0.8Mbps, the rate of load condensate that is the CAN bus has reached 80%, can't satisfy the further requirement of robot multi-freedom control to the communication bandwidth.Industrial Ethernet EPA has the bandwidth of 100Mbps, can support that more transmission node carries out data transmission with higher renewal frequency and instruction sends, overcome the uncertainty that common Ethernet has simultaneously when transmission, guaranteed the reliability and the real-time of bus.
The present invention uses the anthropomorphic robot real-time control method of said system, comprises following two processes:
1, upper motion controller need be when a certain endpoint node sends instruction or data, upper motion controller at first will instruct or data send to the network controller of this endpoint node correspondence by upper layer network, network controller is resolved the packet that receives, extract data division and encapsulate, by lower floor's network the packet that encapsulation obtains is sent to the target endpoint node respectively then according to the form of lower floor's network.
2, endpoint node need be when upper motion controller send state information and feedback information, at first packet is sent to coupled network controller by lower floor's network, network controller resolution data bag content, extract data division and it is encapsulated as the data layout of real-time ethernet, the packet that encapsulation is obtained by upper layer network sends to upper motion controller.
Below in conjunction with the drawings and specific embodiments invention is described in further details.
Fig. 1 is the structural representation of anthropomorphic robot real-time control system provided by the invention.As shown in Figure 1, anthropomorphic robot real-time control system of the present invention comprises a upper motion controller, plurality of network controller and several endpoint nodes.Adopt the double-layer network structure, the plurality of network controller links to each other with upper motion controller by real-time ethernet respectively, forms a upper layer network, carries out communication.Each network controller links to each other with several endpoint nodes respectively; Form lower floor's network, adopt the communication modes that endpoint node adopted to carry out communication.
Upper motion controller is responsible for the planning and the control of robot all-around exercises data, is the main place of carrying out computing.Upper motion controller and network controller carry out communication by real-time ethernet.During communication, the information such as director data that upper motion controller will need to send are packaged into the network controller that the real-time ethernet packet sends to target endpoint node correspondence.
Network controller is responsible for the data communication of upper motion controller of transfer and endpoint node.Network controller is affiliated to simultaneously on real-time ethernet network and lower floor's network.Network controller carries out communication by real-time ethernet and upper motion controller, carries out communication by lower floor's network and several endpoint nodes simultaneously.Each network controller is all connecting lower floor's network, on each lower floor's network a plurality of endpoint nodes is arranged all.When network controller receives the data that upper motion controller sends, earlier packet is resolved, the data content of packet is packaged into the packet that meets lower floor's network call format again, at last new packet is sent on lower floor's network at corresponding target endpoint node place; The real-time network controller receives endpoint node when sending to the data of upper motion controller, equally packet is resolved, and extracts data content and encapsulation, sends to upper motion controller by real-time ethernet at last.Different lower floor's network types that network controller connected can be different, and lower floor's network can be a CAN bus Control Network, 485 bus Control Network, serial line interface network.
Endpoint node is responsible for carrying out the steering order that upper motion controller sends, or the data upload of node is arrived upper motion controller.Endpoint node can be a topworks, as motor driver; Can be sensor also, as pot, plantar pressure sensor, power and torque sensor or attitude sensor.
Fig. 2 is an anthropomorphic robot real-time control system example, by upper motion controller, and three network controllers, a plurality of motor drivers, pot and pressure transducer are formed.Wherein upper motion controller adopts the common pc machine of x86 system, and network controller is the embedded system of ARM system, uses standard ethernet line and hub between upper motion controller and the network controller, connects by the RJ45 network interface.Three network controllers are formed three CAN bus network as lower floor's network with a plurality of endpoint nodes respectively.CAN0 wherein, on two CAN networks of CAN1 respectively carry a plurality of motor drivers, be responsible for the control motor movement; Carry pot and pressure transducer on the CAN2 network are responsible for information acquisition and feedback.
Fig. 3 is an anthropomorphic robot real-time control method step synoptic diagram provided by the invention.As shown in Figure 2, the anthropomorphic robot real-time control method can be divided into data uplink and two processes of data downstream, wherein: data downstream, promptly upper motion controller sends data to endpoint node; Data uplink, promptly endpoint node sends data to upper motion controller.Respectively two steps are specifically described below:
Data downstream step, upper motion controller at first are packaged as a packet with target endpoint node sign, data content waiting for transmission, packet are sent to the network controller of target endpoint node correspondence on the real-time ethernet; Network controller is resolved it after receiving the packet that upper motion controller sends, obtain target endpoint node sign, and the extraction data content, be packaged into the data packet format of lower floor's network correspondence, new packet is sent on lower floor's network at this endpoint node place; Endpoint node receives this packet by lower floor's network, extracts to obtain the data that upper motion controller transmission comes.
Data uplink step, endpoint node are at first with packing data waiting for transmission and send on the lower net network; Network controller is resolved it after receiving packet, extracts data content, is encapsulated as the packet that meets the real-time ethernet form, sends to upper motion controller by real-time ethernet.
Two processes of data uplink and data downstream are independent and can carry out simultaneously.

Claims (5)

1. an anthropomorphic robot real-time control system is characterized in that, comprises a upper motion controller, plurality of network controller and several endpoint nodes etc.; Adopt the double-layer network structure, the plurality of network controller links to each other with upper motion controller by real-time ethernet respectively, forms a upper layer network and carries out communication; Each network controller links to each other with several endpoint nodes respectively; Form lower floor's network and carry out communication.
2. according to the described anthropomorphic robot real-time control system of claim 1, it is characterized in that described real-time ethernet network is Industrial Ethernet EPA(Ethernet for Plant Automation).
3. according to the described anthropomorphic robot real-time control system of claim 1, it is characterized in that described network controller links to each other with several endpoint nodes respectively by CAN bus, 485 buses or serial line interface.
4. according to the described anthropomorphic robot real-time control system of claim 1, it is characterized in that described endpoint node can be motor driver, pot, plantar pressure sensor, power and torque sensor or attitude sensor.
5. an anthropomorphic robot real-time control method of using said system is characterized in that, this method comprises following two processes:
(1) upper motion controller need be when a certain endpoint node sends instruction or data, upper motion controller at first will instruct or data send to the network controller of this endpoint node correspondence by upper layer network, network controller is resolved the packet that receives, extract data division and encapsulate, by lower floor's network the packet that encapsulation obtains is sent to the target endpoint node respectively then according to the form of lower floor's network;
(2) endpoint node need be when upper motion controller send state information and feedback information, at first packet is sent to coupled network controller by lower floor's network, network controller resolution data bag content, extract data division and it is encapsulated as the data layout of real-time ethernet, the packet that encapsulation is obtained by upper layer network sends to upper motion controller.
CN2011101855130A 2011-07-05 2011-07-05 Real-time humanoid robot control system and method Pending CN102298375A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2011101855130A CN102298375A (en) 2011-07-05 2011-07-05 Real-time humanoid robot control system and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2011101855130A CN102298375A (en) 2011-07-05 2011-07-05 Real-time humanoid robot control system and method

Publications (1)

Publication Number Publication Date
CN102298375A true CN102298375A (en) 2011-12-28

Family

ID=45358845

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2011101855130A Pending CN102298375A (en) 2011-07-05 2011-07-05 Real-time humanoid robot control system and method

Country Status (1)

Country Link
CN (1) CN102298375A (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102672719A (en) * 2012-05-10 2012-09-19 浙江大学 Dynamic stability control method for operation of humanoid robot arm
CN103817695A (en) * 2014-02-28 2014-05-28 浙江大学 Control method and drive device for robot flexible joints
CN105847103A (en) * 2016-05-30 2016-08-10 上海理工大学 Method for designing robot multi-channel real-time transmission network with mimic human nervous system
CN105976592A (en) * 2016-07-11 2016-09-28 苏州南江乐博机器人有限公司 Communication device and control method of soccer robot
CN106073896A (en) * 2016-06-15 2016-11-09 重庆金山科技(集团)有限公司 A kind of motor for operating robot controls network and method
CN106444532A (en) * 2016-11-16 2017-02-22 北京航天发射技术研究所 A software and hardware autonomously controllable launching vehicle control system configuration
CN106426195A (en) * 2016-08-31 2017-02-22 佛山博文机器人自动化科技有限公司 Control device and method for humanoid robot
CN106878127A (en) * 2017-01-09 2017-06-20 浙江大学 The wired control system of underwater robot with novel video monitoring

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101373380A (en) * 2008-07-14 2009-02-25 浙江大学 Humanoid robot control system and robot controlling method
CN101907882A (en) * 2010-06-08 2010-12-08 浙江中控电气技术有限公司 Multi-shaft DC servo motor control system and method based on EPA (Ethernet for Plant Automation) field bus

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101373380A (en) * 2008-07-14 2009-02-25 浙江大学 Humanoid robot control system and robot controlling method
CN101907882A (en) * 2010-06-08 2010-12-08 浙江中控电气技术有限公司 Multi-shaft DC servo motor control system and method based on EPA (Ethernet for Plant Automation) field bus

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102672719A (en) * 2012-05-10 2012-09-19 浙江大学 Dynamic stability control method for operation of humanoid robot arm
CN102672719B (en) * 2012-05-10 2014-11-19 浙江大学 Dynamic stability control method for operation of humanoid robot arm
CN103817695A (en) * 2014-02-28 2014-05-28 浙江大学 Control method and drive device for robot flexible joints
CN103817695B (en) * 2014-02-28 2015-10-21 浙江大学 A kind of control method of robot flexibility joint and drive unit
CN105847103A (en) * 2016-05-30 2016-08-10 上海理工大学 Method for designing robot multi-channel real-time transmission network with mimic human nervous system
CN106073896A (en) * 2016-06-15 2016-11-09 重庆金山科技(集团)有限公司 A kind of motor for operating robot controls network and method
CN105976592A (en) * 2016-07-11 2016-09-28 苏州南江乐博机器人有限公司 Communication device and control method of soccer robot
CN106426195A (en) * 2016-08-31 2017-02-22 佛山博文机器人自动化科技有限公司 Control device and method for humanoid robot
CN106444532A (en) * 2016-11-16 2017-02-22 北京航天发射技术研究所 A software and hardware autonomously controllable launching vehicle control system configuration
CN106878127A (en) * 2017-01-09 2017-06-20 浙江大学 The wired control system of underwater robot with novel video monitoring
CN106878127B (en) * 2017-01-09 2019-12-31 浙江大学 Underwater robot wired control system with novel video monitoring function

Similar Documents

Publication Publication Date Title
CN102298375A (en) Real-time humanoid robot control system and method
CN102833112B (en) Communication management machine having looped network function
CN101262473B (en) EPA industrial Ethernet and HART field bus interconnection method
CN103825883A (en) Multi-protocol conversion equipment based on wireless ZigBee, CAN bus and MODBUS/TCP and realization method thereof
CN109639467A (en) Intelligent producing line multi-modal data interactive system and method based on SDN
CN101741766B (en) Conversion device from CAN network to ethernet network
CN102149225A (en) Bullet train state monitoring system and multifunctional hybrid gateway thereof
CN102271100A (en) Gateway device between EtherCAT industrial Ethernet and wireless Zigbee and protocol conversion method of gateway device
CN102082742A (en) Intelligent network protocol conversion and communication control system
CN102332971B (en) Field-bus full duplex reliable communication method for numerical control system
KR102022731B1 (en) Method and gateway for extending ethercat nework
CN104615103A (en) Motor control system and method using ZigBee and CAN bus
CN111083046A (en) 5G-Profibus-DP gateway for industrial field
CN103067238B (en) A kind of collecting method being compatible with MODBUS bus
Li et al. Green and reliable software-defined industrial networks
CN104486783A (en) Multi-state wireless gateway system for multi-state wireless monitoring network and control method
CN103326936A (en) Multi-protocol gateway of Internet of Things allowing unified access of various heterogeneous sensing layer networks
CN105573200A (en) Communication apparatus for wind power converter, and communication method
CN201163782Y (en) Intelligent network gateway used for EPA/HART interconnection
CN103728957B (en) HART (highway addressable remote transducer) field equipment management method and system based on real-time database
CN1216475C (en) Embedded gate for realizing interconnection between networks with different structures
CN102520678A (en) Remote control system for active reflection panel of radio telescope
CN105500372A (en) Modularized welding robot control system based on CAN looped network and control method thereof
CN109120063A (en) A kind of plug & play sensor monitoring method and its system and acquisition unit
CN111413902A (en) Bridge monitoring communication method and acquisition system based on Can bus

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20111228