WO2022242501A1 - 智能机器人的步态控制方法及装置、存储介质、电子装置 - Google Patents

智能机器人的步态控制方法及装置、存储介质、电子装置 Download PDF

Info

Publication number
WO2022242501A1
WO2022242501A1 PCT/CN2022/091877 CN2022091877W WO2022242501A1 WO 2022242501 A1 WO2022242501 A1 WO 2022242501A1 CN 2022091877 W CN2022091877 W CN 2022091877W WO 2022242501 A1 WO2022242501 A1 WO 2022242501A1
Authority
WO
WIPO (PCT)
Prior art keywords
intelligent robot
state information
control
instruction
user instruction
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.)
Ceased
Application number
PCT/CN2022/091877
Other languages
English (en)
French (fr)
Inventor
常新伟
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.)
Dreame Innovation Technology Suzhou Co Ltd
Original Assignee
Dreame Innovation Technology Suzhou Co Ltd
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 Dreame Innovation Technology Suzhou Co Ltd filed Critical Dreame Innovation Technology Suzhou Co Ltd
Publication of WO2022242501A1 publication Critical patent/WO2022242501A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces

Definitions

  • the present invention relates to the communication field, in particular, to a gait control method and device, a storage medium, and an electronic device for an intelligent robot.
  • intelligent robots With the rapid development of intelligent robot technology, intelligent robots have now been applied to various fields in various industries.
  • the existing technology moves the intelligent robot by converting the joint torque of the joint motor into the driving force and driving torque of the robot's center of mass through a dynamic model.
  • the gait control methods of intelligent robots most of them make rigid assumptions about the motion sequence of intelligent robots according to specific motion scenes. User needs and adaptation to real sports scenarios.
  • the motion sequence of the intelligent robot needs to be defined in advance according to specific motion conditions, and the motion sequence of the intelligent robot cannot be dynamically adjusted so that the intelligent robot can adapt to various motion conditions.
  • the purpose of the present invention is to provide a gait control method and device, a storage medium, and an electronic device for an intelligent robot, so as to at least solve the problem in the prior art that the motion sequence of the intelligent robot needs to be defined in advance according to the specific motion situation, and the intelligent robot cannot be dynamically adjusted.
  • Motion sequences enable intelligent robots to adapt to the problem of various motion situations.
  • a gait control method for an intelligent robot including: acquiring a user instruction, wherein the user instruction is used to control at least one of the following: the desired destination of the intelligent robot target position, speed, track point and action sequence; obtain state information, wherein the state information includes: state information of the intelligent robot and state information of the external environment; input the user instruction and the state information into the The whole body dynamics model of the intelligent robot is used to obtain the control instructions of the intelligent robot; and the intelligent robot is controlled according to the control instructions.
  • the method further includes: acquiring control result information after controlling the intelligent robot; adding the control result information to the state information in the collection.
  • the state information of the intelligent robot includes at least one of the following: the first position, the first speed and the first moment of the joint motor of the intelligent robot, and the position and posture information of the intelligent robot;
  • the state information of the environment includes at least one of the following: image information of the external environment, and motion path information generated by the intelligent robot according to the external environment.
  • inputting the user instruction and the state information into the whole-body dynamics model of the intelligent robot to obtain the control instruction of the intelligent robot includes: using the whole-body dynamics model of the intelligent robot to obtain the
  • the successful completion of the user's instruction is the optimization goal, with the foothold position, sole force, foothold sequence and duration of the intelligent robot as variables, the state information as parameters, the whole body dynamics model and foot position,
  • the force on the sole of the foot is a constraint equation, and the optimization parameters are modeled and solved; according to the optimization goal and the motion constraint equation, the control instructions of the intelligent robot are obtained.
  • the motion constraint equation includes:
  • a a1, t ⁇ t
  • t0 ⁇ t ⁇ t1 ⁇ ; a a2, t ⁇ t
  • b b1, t ⁇ t
  • t0 ⁇ t ⁇ t1 ⁇ ; b b2, t ⁇ t
  • q represents the angle of the joint of the intelligent robot
  • dq represents the angular velocity of the joint
  • ddq represents the angular acceleration of the joint
  • G is the gravity
  • F ext and Jc are the external force and other force received by the intelligent robot respectively
  • M is the inertia matrix
  • H is the centripetal force Coriolis force matrix
  • M(q) is the M function about q
  • ⁇ is the joint control torque of the intelligent robot
  • P foot represents the foot end of the intelligent robot
  • the position, t0, t1, t2 are respectively the time point corresponding to the action sequence specified by the user, and a and b are setting parameters.
  • the optimization objective is min(X c -X t ) TR (X c -X t )+ ⁇ T Q ⁇ , where X c and X t are the current state information and Target state information, X c and X t are composed of the intelligent robot's position, velocity, and trajectory points, ⁇ is the joint control torque of the intelligent robot, and R and Q are weight matrices respectively.
  • the method further includes: using the single rigid body model of the intelligent robot, taking the successful completion of the user instruction as the optimization goal, the The optimization goal takes the state information as a parameter, takes the position of the foothold point of the intelligent robot, the force on the sole of the foot, the foothold sequence and the duration as variables, and takes the single rigid body kinematics model, the position of the foot end, and the force on the sole of the foot as the constraint equation, Modeling and solving the optimization parameters; calculate the second position of the joint motor of the intelligent robot through the optimal solution of the position of the foothold point of the intelligent robot, the force on the sole, the sequence and duration of the footfall and the whole body dynamics model of the intelligent robot. Position, second speed and second moment to obtain the control instructions of the intelligent robot.
  • the method further includes: detecting whether the user instruction changes: when it is detected that the user instruction does not change, execute the following steps in a loop until completion At present, the user instruction ends the cycle: obtain the state information, input the user instruction and the state information into the whole body dynamics model of the intelligent robot, obtain the control instruction of the intelligent robot, and control the robot according to the control instruction
  • the intelligent robot is controlled; when it is detected that the user instruction has changed, the following steps are cyclically executed until the current user instruction is completed and the cycle is completed: obtaining the user instruction, obtaining the state information, combining the user instruction and the The state information is input into the whole body dynamics model of the intelligent robot to obtain the control instruction of the intelligent robot, and the intelligent robot is controlled according to the control instruction.
  • a gait control method and device for an intelligent robot including: a first acquiring module, configured to acquire user instructions, wherein the user instructions are used to control at least one of the following : the target position, speed, track point and action sequence to be reached by the intelligent robot; the second acquisition module is used to acquire state information, wherein the state information includes: the state information of the intelligent robot and the external environment State information; a receiving module, configured to input the user instruction and the state information into the whole-body dynamics model of the intelligent robot to obtain a control instruction of the intelligent robot; a control module, configured to control the intelligent robot according to the control instruction The intelligent robot is controlled.
  • a computer-readable storage medium is provided, and a computer program is stored in the storage medium, wherein the computer program is set to execute any of the above-mentioned The gait control method of the intelligent robot described above.
  • an electronic device including a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to perform any of the above The gait control method of the intelligent robot described in item.
  • the present invention has the following beneficial effects: since the user instruction and the state information can be input into the whole body dynamics model of the intelligent robot under the condition of obtaining the user instruction and the state information, the obtained The control instruction of the intelligent robot, and then control the intelligent robot according to the control instruction, so that the intelligent robot can adapt to various motion scenes. Therefore, it solves the problem in the prior art that the motion sequence of the intelligent robot needs to be defined in advance according to the specific motion scene, and the motion sequence of the intelligent robot cannot be dynamically adjusted so that the intelligent robot can adapt to various motion scenes, thereby improving the calculation speed and improving the response of the robot Speed, thereby improving interaction efficiency.
  • Fig. 1 is the block diagram of the hardware structure of the intelligent robot of the gait control method of a kind of intelligent robot of the embodiment of the present invention
  • Fig. 2 is the flowchart of the gait control method of a kind of intelligent robot according to the embodiment of the present invention
  • FIG. 3 is a schematic flow diagram (1) of a gait control method for an intelligent robot according to an embodiment of the present invention
  • FIG. 4 is a schematic flow diagram (2) of a gait control method for an intelligent robot according to an embodiment of the present invention.
  • Fig. 5 is a structural block diagram of a gait control device for an intelligent robot according to an embodiment of the present invention.
  • FIG. 1 is a block diagram of a hardware structure of an intelligent robot according to a gait control method for an intelligent robot according to an embodiment of the present invention.
  • the intelligent robot can include one or more (only one is shown in Figure 1) processor 102 (processor 102 can include but not limited to microprocessor (Microprocessor Unit, MPU for short) or programmable logic Devices (Programmable logic device, PLD for short) and other processing devices and memory 104 for storing data, optionally, the above-mentioned intelligent robot can also include transmission equipment 106 and input and output equipment 108 for communication functions.
  • processor 102 can include but not limited to microprocessor (Microprocessor Unit, MPU for short) or programmable logic Devices (Programmable logic device, PLD for short) and other processing devices and memory 104 for storing data
  • PLD programmable logic device
  • the above-mentioned intelligent robot can also include transmission equipment 106 and input and output equipment 108 for communication functions.
  • the structure shown in Figure 1 is only illustrative, and it does not limit the structure of the above-mentioned intelligent robot.
  • the intelligent robot can also include more or less components than shown in Figure 1, or have the same Functionally equivalent to that shown in Figure 1 or a different configuration with more functionality than shown in Figure 1.
  • the memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the gait control method of the intelligent robot in the embodiment of the present invention, and the processor 102 runs the computer program stored in the memory 104 , so as to perform various functional applications and data processing, that is, to realize the above-mentioned method.
  • the memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory.
  • the memory 104 may further include memory located remotely from the processor 102, and these remote memories may be connected to the intelligent robot through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • the transmission device 106 is used to receive or transmit data via the network.
  • the specific example of the above-mentioned network may include a wireless network provided by the communication provider of the intelligent robot.
  • the transmission device 106 includes a network interface controller (NIC for short), which can be connected to other network devices through a base station so as to communicate with the Internet.
  • the transmission device 106 may be a radio frequency (Radio Frequency, RF for short) module, which is used to communicate with the Internet in a wireless manner.
  • RF Radio Frequency
  • FIG. 2 is a flowchart of a gait control method for an intelligent robot according to an embodiment of the present invention. As shown in FIG. 2 , the process includes the following step:
  • Step S202 acquiring a user instruction, wherein the user instruction is used to control at least one of the following: the target position, speed, track point and action sequence of the intelligent robot to be reached;
  • Step S204 acquiring status information, wherein the status information includes: status information of the intelligent robot and status information of the external environment;
  • Step S206 inputting the user instruction and state information into the whole body dynamics model of the intelligent robot to obtain the control instruction of the intelligent robot;
  • Step S208 controlling the intelligent robot according to the control instruction.
  • the control instruction of the intelligent robot can be obtained, and then The intelligent robot is controlled according to the control instructions, so that the intelligent robot can adapt to various motion scenes. Therefore, it solves the problem in the prior art that the motion sequence of the intelligent robot needs to be defined in advance according to the specific motion scene, and the motion sequence of the intelligent robot cannot be dynamically adjusted so that the intelligent robot can adapt to various motion scenes.
  • the intelligent robot in the embodiment of the present invention may be a biped intelligent robot, a three-legged intelligent robot, a quadruped intelligent robot, or a hexapod intelligent robot, which is not limited in the embodiment of the present invention.
  • step S208 After step S208 is executed, that is, after the intelligent robot is controlled according to the control instruction, the method further includes: acquiring control result information after controlling the intelligent robot; adding the control result information to the In the collection of the state information.
  • step S202 the control result information after the control of the intelligent robot needs to be obtained, and the control result information is added to the set of state information , to update the status information.
  • step S208 the method from step S202 to step S208 is executed in a loop until the current user instruction is completed to end the loop.
  • the state information of the intelligent robot includes at least one of the following: the first position, the first speed and the first moment of the joint motor of the intelligent robot, and the position and posture information of the intelligent robot;
  • the state information of the external environment includes at least one of the following: image information of the external environment, and motion path information generated by the intelligent robot according to the external environment.
  • the state information includes: state information of the intelligent robot and state information of the external environment.
  • the state information of the intelligent robot includes but not limited to: the first position, the first speed and the first moment of the joint motor of the intelligent robot, and the position and posture information of the intelligent robot, wherein the joints of the intelligent robot
  • the first position, the first speed and the first torque of the motor are obtained by measuring the state observer of the intelligent robot, and the position and posture information of the intelligent robot is obtained by measuring the inertial measurement unit.
  • the state information of the external environment includes: image information of the external environment, and movement path information generated by the intelligent robot according to the external environment, wherein the image information of the external environment is obtained by the perception of the intelligent robot Obtained by the module, the perception module can be a camera, an infrared camera module and other equipment used for shooting.
  • the motion path information generated by the intelligent robot according to the external environment is generated by the path planning module of the intelligent robot in actual operation. of.
  • step S206 input the user instruction and the state information into the whole-body dynamics model of the intelligent robot to obtain the control instructions of the intelligent robot, including: through the whole-body dynamics model of the intelligent robot, the
  • the successful completion of the above user instructions is the optimization goal, with the foothold position, sole force, foothold sequence and duration of the intelligent robot as variables, the state information as parameters, and the whole body dynamics model and foot position 1.
  • the force on the sole of the foot is a constraint equation, and the optimization problem is modeled and solved to obtain the most suitable parameters as parameters in the model and equation; according to the optimization target and the motion constraint equation, control instructions for the intelligent robot are obtained.
  • the user instruction and the state information are input into the whole-body dynamics model of the intelligent robot, and an optimization problem is constructed, with the successful completion degree of the user instruction as the optimization goal and the user instruction
  • the highest value of successful completion is the optimization goal, with the successful completion of the user instruction as the optimization goal, with the foothold position, foot force, foothold sequence and duration of the intelligent robot as variables, and with the state information as Parameters, with the whole body dynamic model and the position of the foot end, the force of the sole of the foot as the constraint equation, modeling and solving the optimal parameters.
  • the whole-body dynamics model of the intelligent robot can be solved according to the optimization problem constructed above, through the QP solution method, the foothold position, the force on the sole of the foot, the foothold sequence and the duration of the intelligent robot are solved to obtain the foothold of the intelligent robot.
  • the whole-body dynamics model of the intelligent robot calculates the third position, the third speed and the third moment of the intelligent robot through the solution of the position of the foothold point, the force on the sole of the foot, the sequence of footfall and the duration of the intelligent robot, and then obtains The control instruction of the intelligent robot.
  • the control instruction of the intelligent robot includes the third position, the third speed and the third torque of the joint motor.
  • the third position, the third speed and the third moment are calculated by the whole body dynamics model of the intelligent robot, and the first position, the first speed and the first The moment is measured by the state observer of the intelligent robot.
  • the above-mentioned methods of obtaining the control instructions of the intelligent robot are all calculated in the whole-body dynamics model, wherein, due to the large amount of calculation, the above-mentioned calculation method is difficult to solve with simple constraints, and then substituted into the optimization goal to determine Solution, what the present invention calculates above-mentioned problem adopts is the method for QP solution.
  • the problem in the prior art that the motion sequence of the intelligent robot needs to be defined in advance according to the specific motion conditions and cannot be dynamically adjusted so that the intelligent robot can adapt to various motion conditions is solved.
  • step S206 the method further includes: the motion constraint equation includes:
  • a a1, t ⁇ t
  • t0 ⁇ t ⁇ t1 ⁇ ; a a2, t ⁇ t
  • b b1, t ⁇ t
  • t0 ⁇ t ⁇ t1 ⁇ ; b b2, t ⁇ t
  • q represents the angle of the joint of the intelligent robot
  • dq represents the angular velocity of the joint
  • ddq represents the angular acceleration of the joint
  • G is the gravity
  • F ext and Jc are the external force and other force received by the intelligent robot respectively
  • M is the inertia matrix
  • H is the centripetal force Coriolis force matrix
  • M(q) is the M function about q
  • ⁇ is the joint control torque of the intelligent robot
  • P foot represents the foot end of the intelligent robot
  • the position, t0, t1, t2 are respectively the time point corresponding to the action sequence specified by the user, and a and b are setting parameters.
  • the motion constraint equation is a restriction on the route of the intelligent robot. For example, there are many paths for an intelligent robot to travel from point A to point B, and the motion constraint equation can be understood as the user's constraints on the intelligent robot from point A to point B, for example, it can only go west.
  • the above-mentioned motion constraints include, but are not limited to, friction cone constraints on the soles of the intelligent robot, and terrain restrictions on the body position and posture of the intelligent robot.
  • the motion constraint equation uses the Jacobian matrix, wherein the external force received by the intelligent robot can be converted from the Cartesian space to the joint space through the Jacobian matrix, and then the intelligent robot is controlled to walk through the joint motors of the intelligent robot.
  • the optimization target is, wherein, and are respectively the current state information and the target state information of the intelligent robot, and are composed of the position, speed, and track point of the intelligent robot, and are the intelligent robot's Joint control torques, and are weight matrices, respectively.
  • the current state information and target state information of the intelligent robot are reflected in the execution cycle of the intelligent robot from step S202 to step S208, for example, the intelligent robot obtains the method in the execution cycle of step S202 to step S208
  • the state information is the current state information
  • the target state information is the state information of the intelligent robot corresponding to the fourth position, the fourth speed and the fourth torque of the joint motor in the control instruction of the intelligent robot.
  • the optimization target is the limitation of the solution obtained by the whole-body dynamics model of the intelligent robot according to the motion constraint equation.
  • the whole-body dynamics model of the intelligent robot obtains multiple Solution
  • the optimization goal can be the shortest path among the multiple paths for the intelligent robot to reach from point A to point B, wherein the present invention can use the QP solution method to calculate the above problem.
  • first, second, third, and fourth names in this embodiment are only used to distinguish concepts.
  • first position and the second position may be the same position or different positions. The same is true for different speeds and different torques, so I won't repeat them here.
  • the method described above is to obtain the control instructions of the intelligent robot completely through the whole body dynamics model, which is the first method adopted by the present invention, and the present invention also provides a second method: through a single Rigid body model and whole body dynamics model obtain the method for the control instruction of described intelligent robot:
  • the successful completion of the user instruction is used as the optimization goal
  • the optimization goal takes the state information as a parameter
  • the optimization goal takes the state information as a parameter
  • the position of the foothold point of the intelligent robot, the force on the soles of the feet, the foothold sequence and the duration are variables
  • the single rigid body kinematics model, the position of the foot end, and the force on the soles of the feet are used as the constraint equations to model and solve the optimal parameters to obtain the intelligent robot.
  • the position of the foothold point, the force on the sole of the foot, the solution of the foothold sequence and the duration and then input the solution of the foothold point position, the force on the sole of the foot, the footstep sequence and the duration of the intelligent robot into the whole body dynamics model of the intelligent robot, and calculate the The second position, the second speed and the second torque of the joint motors of the intelligent robot are obtained, and then the control instructions of the intelligent robot are obtained.
  • the solution of the position of the foothold, the force on the sole of the foot, the footfall sequence and the duration of the intelligent robot is first obtained through the single rigid body model, and then the joint motors of the intelligent robot are calculated according to the whole-body dynamics model.
  • Second position, second speed and second torque In the first method, the whole-body dynamics model first obtains the solution of the position of the foothold, the force on the sole of the foot, the sequence of footfall and the duration of the intelligent robot, and then calculates the fourth position, the fourth position, and the duration of the joint motor of the intelligent robot. Fourth speed and fourth torque.
  • the method of obtaining the control instructions of the intelligent robot through the whole-body dynamics model since only the whole-body dynamics model is involved in the whole process, it can be directly regarded as that the whole-body dynamics model is obtained once The position of the foothold of the intelligent robot, the force on the sole of the foot, the solution of the foothold sequence and duration, and the position, speed and torque of the joint motor.
  • the method further includes: detecting whether the user instruction changes: when it is detected that the user instruction does not change, execute the following steps in a loop until the current user instruction is completed.
  • Instruction end cycle obtain the state information, input the user instruction and the state information into the whole body dynamics model of the intelligent robot, obtain the control instruction of the intelligent robot, and control the intelligent robot according to the control instruction Control; when it is detected that the user instruction changes, execute the following steps in a loop until the current user instruction is completed and end the cycle: obtain the user instruction, obtain the state information, and input the user instruction and the state information
  • the whole body dynamics model of the intelligent robot obtains the control instruction of the intelligent robot, and controls the intelligent robot according to the control instruction.
  • loop execution obtain the state information, input the user instruction and the state information into the whole body dynamics model of the intelligent robot, obtain the control instruction of the intelligent robot, The intelligent robot is controlled according to the control instructions. End the loop until the current user instruction is completed. If it is detected that the user instruction has changed, execute in a loop: obtain the user instruction, obtain the state information, input the user instruction and the state information into the whole body dynamics model of the intelligent robot, and obtain the intelligent robot A control instruction of the robot, and the intelligent robot is controlled according to the control instruction. The loop is ended until the current user instruction is completed.
  • the present invention also provides an optional embodiment for explaining the above technical solution.
  • Fig. 3 is the schematic flow chart (1) of the gait control method of a kind of intelligent robot according to the embodiment of the present invention. As shown in Fig. 3:
  • S302 Obtain user instructions, wherein the user instructions are used to control at least one of the following: the target position, speed, track point and action sequence to be reached by the intelligent robot, wherein the action sequence includes but is not limited to arriving at a specified time Complete the specified action at the specified location;
  • S304 Obtain state information through a state observer, wherein the state information includes: the state information of the intelligent robot and the state information of the external environment, the first position, the first speed, and the first position of the joint motor of the intelligent robot.
  • the state information of the external environment includes: the image information of the external environment, and the motion path information generated by the intelligent robot according to the external environment, wherein the path planning module The generated motion path information is obtained through the perception module to obtain image information of the external environment;
  • S308 Control the joints of the intelligent robot through the fourth position, the fourth speed and the fourth torque of the joint motor.
  • FIG. 4 is a schematic flow diagram (2) of a gait control method for an intelligent robot according to an embodiment of the present invention.
  • S402 Obtain user instructions, wherein the user instructions are used to control at least one of the following: the target position, speed, track point and action sequence to be reached by the intelligent robot, wherein the action sequence includes but is not limited to arriving at a specified time Complete the specified action at the specified location;
  • S404 Obtain state information through a state observer, wherein the state information includes: the state information of the intelligent robot and the state information of the external environment, the first position, the first speed, and the first position of the joint motor of the intelligent robot.
  • the state information of the external environment includes: the image information of the external environment, and the motion path information generated by the intelligent robot according to the external environment, wherein the path planning module The generated motion path information is obtained through the perception module to obtain image information of the external environment;
  • S410 Control the joints of the intelligent robot through the fourth position, the fourth speed and the fourth torque of the joint motor.
  • the control instruction of the intelligent robot can be obtained, and then The intelligent robot is controlled according to the control instructions, so that the intelligent robot can adapt to various motion scenes. Therefore, it solves the problem in the prior art that the motion sequence of the intelligent robot needs to be defined in advance according to the specific motion scene, and the motion sequence of the intelligent robot cannot be dynamically adjusted so that the intelligent robot can adapt to various motion scenes.
  • the technical solution of the present application can be embodied in the form of a software product in essence or the part that contributes to the prior art, and the computer software product is stored in a storage medium (such as a read-only memory (Read-Only Memory, abbreviated as ROM), random access memory (Random Access Memory, abbreviated as RAM), magnetic disk, optical disk), including several instructions to make a terminal device (which can be a mobile phone, computer, server, or network device) etc.) to implement the methods described in the various embodiments of the present application.
  • a storage medium such as a read-only memory (Read-Only Memory, abbreviated as ROM), random access memory (Random Access Memory, abbreviated as RAM), magnetic disk, optical disk
  • a terminal device which can be a mobile phone, computer, server, or network device
  • a gait control device for an intelligent robot is also provided, and the gait control device for an intelligent robot is used to implement the above-mentioned embodiments and preferred implementation modes, and those that have already been described will not be repeated.
  • the term "module” may be a combination of software and/or hardware that realizes a predetermined function.
  • the devices described in the following embodiments are preferably implemented in software, implementations in hardware, or a combination of software and hardware are also possible and contemplated.
  • Fig. 5 is the structural block diagram of the gait control device of a kind of intelligent robot according to the embodiment of the present invention, as shown in Fig. 5:
  • the first obtaining module 50 is used to obtain user instructions, wherein the user instructions are used to control at least one of the following: the target position, speed, track point and action sequence to be reached by the intelligent robot;
  • the second acquisition module 52 is used to acquire status information, wherein the status information includes: status information of the intelligent robot and status information of the external environment;
  • the receiving module 54 is used to input user instructions and state information into the whole body dynamics model of the intelligent robot to obtain control instructions of the intelligent robot;
  • the control module 56 is used to control the intelligent robot according to the control instructions.
  • the control instruction of the intelligent robot can be obtained, and then The intelligent robot is controlled according to the control instructions, so that the intelligent robot can adapt to various motion scenes. Therefore, it solves the problem in the prior art that the motion sequence of the intelligent robot needs to be defined in advance according to the specific motion scene, and the motion sequence of the intelligent robot cannot be dynamically adjusted so that the intelligent robot can adapt to various motion scenes.
  • the intelligent robot in the embodiment of the present invention may be a biped intelligent robot, a three-legged intelligent robot, a quadruped intelligent robot, or a hexapod intelligent robot, which is not limited in the embodiment of the present invention.
  • control module 56 is further configured to acquire control result information after controlling the intelligent robot; add the control result information to the set of state information.
  • steps S202 to S208 are executed in a loop until the current user instruction is completed to end the loop.
  • the second acquisition module 52 is also used for the first position, the first speed and the first moment of the joint motor of the intelligent robot, and the position and posture information of the intelligent robot;
  • the state information of the external environment includes : image information of the external environment, and motion path information generated by the intelligent robot according to the external environment.
  • the state information includes: state information of the intelligent robot and state information of the external environment.
  • the state information of the intelligent robot includes but not limited to: the first position, the first speed and the first moment of the joint motor of the intelligent robot, and the position and posture information of the intelligent robot, wherein the joints of the intelligent robot
  • the first position, the first speed and the first torque of the motor are obtained by measuring the state observer of the intelligent robot, and the position and posture information of the intelligent robot is obtained by measuring the inertial measurement unit.
  • the state information of the external environment includes: image information of the external environment, and movement path information generated by the intelligent robot according to the external environment, wherein the image information of the external environment is obtained by the perception of the intelligent robot Obtained by the module, the perception module can be a camera, an infrared camera module and other equipment used for shooting.
  • the motion path information generated by the intelligent robot according to the external environment is generated by the path planning module of the intelligent robot in actual operation. of.
  • the receiving module 54 is also used to use the whole-body dynamics model of the intelligent robot to take the successful completion of the user's instruction as the optimization goal, and take the position of the foothold, the force on the sole of the foot, and the foothold of the intelligent robot as the optimization goal.
  • the sequence and duration are variables, and the state information is used as a parameter to establish a motion constraint equation; according to the optimization goal and the motion constraint equation, control instructions for the intelligent robot are obtained.
  • the user instruction and the state information are input into the whole-body dynamics model of the intelligent robot to construct an optimization problem, with the successful completion of the user instruction as the optimization goal, and the state information
  • the whole-body dynamics model of the intelligent robot solves the position of the foothold point, the force on the sole of the foot, the foothold sequence and the duration of the intelligent robot according to the optimization problem constructed above, and obtains the foothold position of the intelligent robot , plantar force, landing sequence and duration solution.
  • the whole body dynamics model of the intelligent robot calculates the second position, the second speed and the second moment of the intelligent robot through the solution of the position of the foothold point of the intelligent robot, the force on the sole of the foot, the footfall sequence and the duration, and then obtains The control instruction of the intelligent robot.
  • the control command of the intelligent robot includes the fourth position, the fourth speed and the fourth torque of the joint motor.
  • the fourth position, the fourth speed and the fourth moment are calculated by the whole body dynamics model of the intelligent robot, and the first position, the first speed and the first The moment is measured by the state observer of the intelligent robot.
  • the above-mentioned methods for obtaining the control commands of the intelligent robot are all calculated in the whole-body dynamics model, where, due to the large amount of calculation, it is difficult to solve the above-mentioned calculation method with simple constraints, and then substitute into Optimizing the objective to determine the solution, the present invention calculates the above-mentioned problem and adopts the method of QP solution.
  • the problem in the prior art that the motion sequence of the intelligent robot needs to be defined in advance according to the specific motion conditions and cannot be dynamically adjusted so that the intelligent robot can adapt to various motion conditions is solved.
  • the motion constraint equation described in the receiving module 54 includes:
  • a a1, t ⁇ t
  • t0 ⁇ t ⁇ t1 ⁇ ; a a2, t ⁇ t
  • b b1, t ⁇ t
  • t0 ⁇ t ⁇ t1 ⁇ ; b b2, t ⁇ t
  • q represents the angle of the joint of the intelligent robot
  • dq represents the angular velocity of the joint
  • ddq represents the angular acceleration of the joint
  • G is the gravity
  • F ext and Jc are the external force and other force received by the intelligent robot respectively
  • M is the inertia matrix
  • H is the centripetal force Coriolis force matrix
  • M(q) is the M function about q
  • ⁇ is the joint control torque of the intelligent robot
  • P foot represents the foot end of the intelligent robot
  • the position, t0, t1, t2 are respectively the time point corresponding to the action sequence specified by the user, and a and b are setting parameters.
  • the motion constraint equation is a restriction on the travel route of the intelligent robot.
  • the motion constraint equation can be understood as the user's constraints on the intelligent robot from point A to point B, for example, it can only go west.
  • the above-mentioned motion constraints include but are not limited to the friction cone constraint of the intelligent robot's foot force, and the terrain's restrictions on the position and posture of the intelligent robot's fuselage.
  • the motion constraint equation uses the Jacobian matrix, wherein the external force received by the intelligent robot can be converted from the Cartesian space to the joint space through the Jacobian matrix, and then the intelligent robot is controlled to walk through the joint motors of the intelligent robot.
  • the optimization target in the receiving module 54 is min(X c -X t ) TR (X c -X t )+ ⁇ T Q ⁇ , where X c and X t are the Current state information and target state information, X c and X t are composed of the intelligent robot's position, speed, and trajectory points, ⁇ is the joint control torque of the intelligent robot, and R and Q are weight matrices respectively.
  • the state information acquired by the intelligent robot is current state information
  • the target state information is the state information of the intelligent robot corresponding to the position, speed and torque of the joint motor in the control command of the intelligent robot.
  • the optimization target is the limitation of the solution obtained by the whole-body dynamics model of the intelligent robot according to the motion constraint equation.
  • the whole-body dynamics model of the intelligent robot obtains multiple Solution
  • the optimization goal can be the shortest path among the multiple paths for the intelligent robot to reach from point A to point B, wherein the present invention can use the QP solution method to calculate the above problem.
  • the receiving module 54 is also configured to use the single rigid body model of the intelligent robot, take the successful completion of the user instruction as the optimization goal, and use the state information as a parameter to determine the foothold of the intelligent robot Position, foot force, foot sequence and duration are solved to obtain the solution of the foothold position, foot force, foot sequence and duration of the intelligent robot; through the foothold position, foot force, and foot sequence of the intelligent robot Calculate the second position, the second speed and the second torque of the joint motors of the intelligent robot based on the solution of the time length and the whole body dynamics model of the intelligent robot, so as to obtain the control instructions of the intelligent robot.
  • control instructions of the intelligent robot can be obtained completely through the whole-body dynamics model, and the present invention also provides a method for obtaining the control instructions of the intelligent robot through the single rigid body model and the whole-body dynamics model.
  • the successful completion degree of the user instruction is used as the optimization goal, and the state information is used as a parameter.
  • the location of the foothold, the force on the sole, the sequence of footfall and the duration are solved to obtain the solution of the position of the foothold, the force on the sole of the foot, the sequence of foothold and the duration of the intelligent robot.
  • input the solution of the foothold position of the intelligent robot, the force on the soles of the feet, the footfall sequence and the duration into the whole-body dynamics model of the intelligent robot, and calculate the second position, the second speed and the second speed of the joint motors of the intelligent robot.
  • the second torque and then obtain the control instruction of the intelligent robot.
  • the receiving module 54 is also used to detect whether the user instruction has changed: when it is detected that the user instruction remains unchanged, the following steps are executed in a loop until the current user instruction is completed and the cycle is ended: the state information is obtained, and the The user instruction and the state information are input into the whole body dynamics model of the intelligent robot to obtain the control instruction of the intelligent robot, and the intelligent robot is controlled according to the control instruction; when it is detected that the user instruction occurs change, the following steps are executed cyclically until the current user instruction is completed and the cycle is ended: obtain the user instruction, obtain the state information, input the user instruction and the state information into the whole body dynamics model of the intelligent robot, and obtain The control instruction of the intelligent robot is used to control the intelligent robot according to the control instruction.
  • loop execution obtain the state information, input the user instruction and the state information into the whole body dynamics model of the intelligent robot, obtain the control instruction of the intelligent robot, The intelligent robot is controlled according to the control instruction. End the loop until the current user instruction is completed. If it is detected that the user instruction has changed, execute in a loop: obtain the user instruction, obtain the state information, input the user instruction and the state information into the whole body dynamics model of the intelligent robot, and obtain the intelligent robot A control instruction of the robot, and the intelligent robot is controlled according to the control instruction. End the loop until the currently described user command is completed
  • the above-mentioned modules can be realized by software or hardware. For the latter, it can be realized by the following methods, but not limited to this: the above-mentioned modules are all located in the same processor; or, the above-mentioned modules can be combined in any combination The forms of are located in different processors.
  • Embodiments of the present invention also provide a computer-readable storage medium, in which a computer program is stored, wherein the computer program is set to execute the steps in any one of the above method embodiments when running.
  • the above-mentioned storage medium may be configured to store a computer program for performing the following steps: S1. Obtain a user instruction, wherein the user instruction is used to control at least one of the following: the intelligence The target position, speed, track point and action sequence of the robot to be reached;
  • status information includes: status information of the intelligent robot and status information of the external environment;
  • the above-mentioned storage medium may include but not limited to: various media capable of storing computer programs such as USB flash drive, read-only memory ROM, random access memory RAM, mobile hard disk, magnetic disk or optical disk.
  • various media capable of storing computer programs such as USB flash drive, read-only memory ROM, random access memory RAM, mobile hard disk, magnetic disk or optical disk.
  • An embodiment of the present invention also provides an electronic device, including a memory and a processor, where a computer program is stored in the memory, and the processor is configured to run the computer program to perform the steps in any one of the above method embodiments.
  • the above-mentioned electronic device may further include a transmission device and an input-output device, wherein the transmission device is connected to the above-mentioned processor, and the input-output device is connected to the above-mentioned processor.
  • the above-mentioned processor may be configured to execute the following steps through a computer program:
  • status information includes: status information of the intelligent robot and status information of the external environment;
  • each module or each step of the above-mentioned present invention can be realized by a general-purpose computing device, and they can be concentrated on a single computing device, or distributed in a network formed by multiple computing devices Alternatively, they may be implemented in program code executable by a computing device so that they may be stored in a storage device to be executed by a computing device, and in some cases in an order different from that shown here
  • the steps shown or described are carried out, or they are separately fabricated into individual integrated circuit modules, or multiple modules or steps among them are fabricated into a single integrated circuit module for implementation.
  • the present invention is not limited to any specific combination of hardware and software.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Manipulator (AREA)

Abstract

本发明提供了一种智能机器人的步态控制方法及装置、存储介质、电子装置，上述方法包括：获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息；将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令；根据所述控制指令对所述智能机器人进行控制。采用上述技术方案，解决现有技术中需要根据具体运动场景提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动场景的问题。

Description

智能机器人的步态控制方法及装置、存储介质、电子装置
本公开要求如下专利申请的优先权：于2021年05月18提交中国专利局、申请号为202110537158.2、发明名称为“智能机器人的步态控制方法及装置、存储介质、电子装置”的中国专利申请；于2021年05月18提交中国专利局、申请号为202111192823.5、发明名称为“智能机器人的步态控制方法及装置、存储介质、电子装置”的中国专利申请；上述专利申请的全部内容通过引用结合在本公开中。
技术领域
本发明涉及通信领域，具体而言，涉及一种智能机器人的步态控制方法及装置、存储介质、电子装置。
背景技术
随着智能机器人技术的快速发展，智能机器人现在已经被应用到了各行各业的各个领域。现有技术通过将关节电机的关节力矩经过动力学模型转化为机器人质心的驱动力和驱动力矩来移动智能机器人。目前，智能机器人步态控制方法中，大多根据具体运动场景对智能机器人的运动序列做出硬性假设，这些方法只适用于智能机器人特定运动情况和特定环境，难以满足智能机器人实际使用中多样化的用户需求和适应真实运动场景。
现有技术存在需要根据具体运动情况提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动情况的问题。
因此，有必要对现有技术予以改良以克服现有技术中的所述缺陷。
发明内容
本发明的目的在于提供一种智能机器人的步态控制方法及装置、存储介质、电子装置，以至少解决现有技术中需要根据具体运动情况提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动情况的问题。
本发明的目的是通过以下技术方案实现：
根据本发明的一个可选实施例，提供了一种智能机器人的步态控制方法，包括：获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；获取状态信息，其中，所述状态信息包括：所述智能机器人 的状态信息和外部环境的状态信息；将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令；根据所述控制指令对所述智能机器人进行控制。
可选的，根据所述控制指令对所述智能机器人进行控制之后，所述方法还包括：获取对所述智能机器人控制之后的控制结果信息；将所述控制结果信息添加到给所述状态信息的集合中。
可选的，所述智能机器人的状态信息包括至少以下之一：所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息；所述外部环境的状态信息包括至少以下之一：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息。
可选的，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，包括：通过所述智能机器人的全身动力学模型，以所述用户指令的成功完成度为最优化目标，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以所述状态信息为参数，以全身动力学模型和足端位置、脚底受力为约束方程，建模求解最优化参数；根据最优化目标和所述运动约束方程，得到所述智能机器人的控制指令。
可选的，所述运动约束方程，包括：
等式约束1：M(q)ddq+H(q,dq)+G-J cF ext＝τ
等式约束2：P foot(t)＝a
a＝a1，t∈{t|t0＜t＜t1}；a＝a2，t∈{t|t1＜t＜t2}
不等式约束：F ext(t)＞b
b＝b1，t∈{t|t0＜t＜t1}；b＝b2，t∈{t|t1＜t＜t2}
其中，q代表所述智能机器人的关节的角度，dq代表所述关节的角速度，ddq代表所述关节的角加速度，G为重力，F ext和J c分别是所述智能机器人受到的外力和其对应的雅克比矩阵，M是惯性矩阵，H是向心力科氏力矩阵，M(q)是关于q的M函数，τ为所述智能机器人的关节控制力矩，P foot代表所述智能机器人足端位置，t0、t1、t2分别为与用户指定的动作序列所对应的时刻点，a和b是设置参数。
可选的，所述最优化目标为min(X c-X t) TR(X c-X t)+τ TQτ，其中，X c和X t分别为所述智能机器人的当前状态信息和目标状态信息，X c和X t由所述智能机器人位置、速度、轨 迹点所组成，τ为所述智能机器人的关节控制力矩，R和Q分别为权重矩阵。
可选的，在根据所述控制指令对所述智能机器人进行控制之前，所述方法还包括：通过所述智能机器人的单刚体模型，以所述用户指令的成功完成度为最优化目标，该最优化目标以所述状态信息为参数，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以单刚体运动学模型和足端位置、脚底受力为约束方程，建模求解最优化参数；通过所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的最优解和所述智能机器人的全身动力学模型计算所述智能机器人的关节电机的第二位置、第二速度和第二力矩，以得到所述智能机器人的控制指令。
可选的，在根据所述控制指令对所述智能机器人进行控制之后，所述方法还包括：检测所述用户指令是否发生变化：当检测到所述用户指令不变，循环执行以下步骤直到完成当前所述用户指令结束循环：获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制；当检测到所述用户指令发生变化，循环执行以下步骤直至完成当前所述用户指令结束循环：获取所述用户指令，获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。
根据本发明的又一个可选实施例，提供了一种智能机器人的步态控制方法装置，包括：第一获取模块，用于获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；第二获取模块，用于获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息；接收模块，用于将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令；控制模块，用于根据所述控制指令对所述智能机器人进行控制。
根据本发明的又一个可选实施例，提供了一种计算机可读的存储介质，所述存储介质中存储有计算机程序，其中，所述计算机程序被设置为运行时执行以上任一项中所述的智能机器人的步态控制方法。
根据本发明的又一个可选实施例，提供了一种电子装置，包括存储器和处理器，所述存储器中存储有计算机程序，所述处理器被设置为运行所述计算机程序以执行以上任一项中所述的智能机器人的步态控制方法。
与现有技术相比，本发明具有如下有益效果：由于可以在获取用户指令和获取状态信息的情况下，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，进而根据所述控制指令对所述智能机器人进行控制，从而使得智 能机器人适应各种运动场景。因此，解决了现有技术中需要根据具体运动场景提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动场景的问题，因而能够提高计算速度，提高机器人的反应速度，进而提高交互效率。
附图说明
此处所说明的附图用来提供对本发明的进一步理解，构成本发明的一部分，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的不当限定。在附图中：
图1是本发明实施例的一种智能机器人的步态控制方法的智能机器人的硬件结构框图；
图2是根据本发明实施例的一种智能机器人的步态控制方法的流程图；
图3是根据本发明实施例的一种智能机器人的步态控制方法的流程示意图(一)；
图4是根据本发明实施例的一种智能机器人的步态控制方法的流程示意图(二)；
图5为根据本发明实施例的一种智能机器人的步态控制装置的结构框图。
具体实施方式
下文中将参考附图并结合实施例来详细说明本发明。需要说明的是，在不冲突的情况下，本发明中的实施例及实施例中的特征可以相互组合。
需要说明的是，本发明的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。
本发明实施例所提供的方法实施例可以在智能机器人或者类似的运算装置中执行。以运行在智能机器人上为例，图1是本发明实施例的一种智能机器人的步态控制方法的智能机器人的硬件结构框图。如图1所示，智能机器人可以包括一个或多个(图1中仅示出一个)处理器102(处理器102可以包括但不限于微处理器(Microprocessor Unit,简称是MPU)或可编程逻辑器件(Programmable logic device,简称是PLD)等的处理装置和用于存储数据的存储器104，可选地，上述智能机器人还可以包括用于通信功能的传输设备106以及输入输出设备108。本领域普通技术人员可以理解，图1所示的结构仅为示意，其并不对上述智能机器人的结构造成限定。例如，智能机器人还可包括比图1中所示更多或者更少的组件，或者具有与图1所示等同功能或比图1所示功能更多的不同的配置。
存储器104可用于存储计算机程序，例如，应用软件的软件程序以及模块，如本发明实施例中的智能机器人的步态控制方法对应的计算机程序，处理器102通过运行存储在存储器104内的计算机程序，从而执行各种功能应用以及数据处理，即实现上述的方法。存储器104可包括高速随机存储器，还可包括非易失性存储器，如一个或者多个磁性存储装置、闪存、或 者其他非易失性固态存储器。在一些实例中，存储器104可进一步包括相对于处理器102远程设置的存储器，这些远程存储器可以通过网络连接至智能机器人。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
传输设备106用于经由网络接收或者发送数据。上述的网络具体实例可包括智能机器人的通信供应商提供的无线网络。在一个实例中，传输设备106包括一个网络适配器(Network Interface Controller，简称为NIC)，其可通过基站与其他网络设备相连从而可与互联网进行通讯。在一个实例中，传输设备106可以为射频(Radio Frequency，简称为RF)模块，其用于通过无线方式与互联网进行通讯。
在本实施例中提供了一种运行于上述智能机器人的步态控制方法，图2是根据本发明实施例的智能机器人的步态控制方法的流程图，如图2所示，该流程包括如下步骤：
步骤S202，获取用户指令，其中，用户指令用于控制至少以下之一：智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；
步骤S204，获取状态信息，其中，状态信息包括：智能机器人的状态信息和外部环境的状态信息；
步骤S206，将用户指令和状态信息输入智能机器人的全身动力学模型，得到智能机器人的控制指令；
步骤S208，根据控制指令对智能机器人进行控制。
通过上述技术方案，由于可以在获取用户指令和获取状态信息的情况下，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，进而根据所述控制指令对所述智能机器人进行控制，从而使得智能机器人适应各种运动场景。因此，解决了现有技术中需要根据具体运动场景提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动场景的问题。
需要说明的是，本发明实施例的智能机器人可以是双足智能机器人，三足智能机器人，还可以是四足智能机器人、六足智能机器人，本发明实施例对此不进行限定。
在执行步骤S208之后，也就是根据所述控制指令对所述智能机器人进行控制之后，所述方法还包括：获取对所述智能机器人控制之后的控制结果信息；将所述控制结果信息添加到给所述状态信息的集合中。
需要说明的是，在根据所述控制指令对所述智能机器人进行控制之后，需要获取对所述智能机器人控制之后的控制结果信息，并将所述控制结果信息添加到所述状态信息的集合中，以更新所述状态信息。在更新所述状态信息之后，循环执行步骤S202至步骤S208的方法，直到完成当前所述用户指令结束循环。
在步骤S204中，所述智能机器人的状态信息包括至少以下之一：所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息；所述外部环境的状态信息包括至少以下之一：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息。
需要说明的是，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息。所述智能机器人的状态信息包括但不限于：所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息，其中，所述智能机器人的关节电机的第一位置、第一速度和第一力矩是由所述智能机器人的状态观测器测量得到的，所述智能机器人的位置姿势信息是惯性测量单元测量得到的。所述外部环境的状态信息包括：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息，其中，所述外部环境的图像信息是由所述智能机器人的感知模块获取的，感知模块可以是相机，红外摄像模组等用于拍摄的设备，所述智能机器人根据所述外部环境生成的运动路径信息，在实际操作中是所述智能机器人的路径规划模块生成的。
在步骤S206中，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，包括：通过所述智能机器人的全身动力学模型，以所述用户指令的成功完成度为最优化目标，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以所述状态信息为参数，以全身动力学模型和足端位置、脚底受力为约束方程，建模求解最优化问题，以得到最合适的参数作为模型和方程中的参数；根据最优化目标和所述运动约束方程，得到所述智能机器人的控制指令。
需要说明的是，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，构建最优化问题，以所述用户指令的成功完成度为最优化目标以所述用户指令的成功完成度的最高值为优化目标以所述用户指令的成功完成度为最优化目标，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以所述状态信息为参数，以全身动力学模型和足端位置、脚底受力为约束方程，建模求解最优化参数。可以所述智能机器人的全身动力学模型根据上述构建的最优化问题，通过QP求解方法对所述智能机器人的落脚点位置、脚底受力、落脚序列和时长进行求解，得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解。所述智能机器人的全身动力学模型再通过所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解计算所述智能机器人的第三位置、第三速度和第三力矩，进而得到所述智能机器人的控制指令。其中，所述智能机器人的控制指令包括所述关节电机的第三位置、第三速度和第三力矩。需要说明的是，所述第三位置、第三速度和第三力矩是所述智能机器人的全身动力学模型计算出来的，所述智能机器人的关节电机的第一位置、第一速度 和第一力矩是由所述智能机器人的状态观测器测量得到的。
上述得到所述智能机器人的控制指令的方法都是在所述全身动力学模型中计算的，其中，由于计算量太大，上述计算方法很难以简单的约束条件求解，然后代入最优化目标来确定解，本发明计算上述问题采用的是QP求解的方法。通过上述技术手段，解决了现有技术存在需要根据具体运动情况提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动情况的问题。
步骤S206中，所述方法还包括：所述运动约束方程，包括：
等式约束1：M(q)ddq+H(q,dq)+G-J cF ext＝τ
等式约束2：P foot(t)＝a
a＝a1，t∈{t|t0＜t＜t1}；a＝a2，t∈{t|t1＜t＜t2}
不等式约束：F ext(t)＞b
b＝b1，t∈{t|t0＜t＜t1}；b＝b2，t∈{t|t1＜t＜t2}
其中，q代表所述智能机器人的关节的角度，dq代表所述关节的角速度，ddq代表所述关节的角加速度，G为重力，F ext和J c分别是所述智能机器人受到的外力和其对应的雅克比矩阵，M是惯性矩阵，H是向心力科氏力矩阵，M(q)是关于q的M函数，τ为所述智能机器人的关节控制力矩，P foot代表所述智能机器人足端位置，t0、t1、t2分别为与用户指定的动作序列所对应的时刻点，a和b是设置参数。
所述运动约束方程是对智能机器人的行进路线的限制。举例说明，智能机器人从A地到达B地，有很多条路径，所述运动约束方程就可以理解是使用者对智能机器人从A地到达B地，比如只能向西走等约束条件。其中，上述运动约束包括但不限于智能机器人脚底受力的摩擦锥约束、地形对智能机器人的机身位置姿势的限制。所述运动约束方程用到了雅克比矩阵，其中通过雅克比矩阵就可以把智能机器人受到的外力从笛卡尔空间转换到关节空间，进而，通过智能机器人的关节电机控制智能机器人行走。
步骤S206中，所述最优化目标为，其中，和分别为所述智能机器人的当前状态信息和目标状态信息，和由所述智能机器人位置、速度、轨迹点所组成，为所述智能机器人的关节控制力矩，和分别为权重矩阵。
所述智能机器人的当前状态信息和目标状态信息，是所述智能机器人在执行循环执行步骤S202至步骤S208中体现的，比如，所述智能机器人在执行循环执行步骤S202至步骤S208的方法中获取的状态信息是当前状态信息，目标状态信息是所述智能机器人的控制指令中所述关节电机的第四位置、第四速度和第四力矩对应的所述智能机器人的状态信息。所述最优 化目标是所述智能机器人的全身动力学模型根据所述运动约束方程求出的解的限制，比如，所述智能机器人的全身动力学模型根据所述运动约束方程求出了多个解，所述最优化目标可以是智能机器人从A地到达B地的多个路径中最短的路径，其中，本发明计算上述问题可以采用QP求解的方法。
需要说明的是，本实施例中第一、第二、第三、第四等名称仅为用于概念区分，例如第一位置和第二位置可能是同一个位置，也可能是不同位置，对应的不同速度和不同力矩也一样，不再赘述。
需要说明的是，上述记载的方法是完全通过所述全身动力学模型得到所述智能机器人的控制指令，是本发明所采用的第一种方法，本发明还提供了第二种方法：通过单刚体模型和全身动力学模型得到所述智能机器人的控制指令的方法：
首先通过所述智能机器人的单刚体模型，以所述用户指令的成功完成度为最优化目标，该最优化目标以所述状态信息为参数，该最优化目标以所述状态信息为参数，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以单刚体运动学模型和足端位置、脚底受力为约束方程，建模求解最优化参数，得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解，然后将所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解输入所述智能机器人的全身动力学模型，计算所述智能机器人的关节电机的第二位置、第二速度和第二力矩，进而得到所述智能机器人的控制指令。
本实施例中，是先通过所述单刚体模型得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解，然后根据全身动力学模型，计算所述智能机器人的关节电机的第二位置、第二速度和第二力矩。而第一种方法中，是所述全身动力学模型先得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解，然后计算所述智能机器人的关节电机的第四位置、第四速度和第四力矩。
需要说明的是，通过所述全身动力学模型得到所述智能机器人的控制指令的方法，由于整个过程中只涉及所述全身动力学模型，所以可以直接看成所述全身动力学模型一次就得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解，所述关节电机的位置、速度和力矩。
在根据所述控制指令对所述智能机器人进行控制之后，所述方法还包括：检测所述用户指令是否发生变化：当检测到所述用户指令不变，循环执行以下步骤直到完成当前所述用户指令结束循环：获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制；当检测到所述用户指令发生变化，循环执行以下步骤直至完成当前所述用户指令结束 循环：获取所述用户指令，获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。
需要说明的是，在根据所述控制指令对所述智能机器人进行控制之后，还需要检测所述用户指令是否发生变化。如果检测到所述用户指令不变，那么循环执行：获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。直到完成当前所述用户指令结束循环。如果检测到所述用户指令发生变化，循环执行：获取所述用户指令，获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。直至完成当前所述用户指令结束循环。
为了更好的理解上述技术方案，本发明还提供了一种可选实施例，用于解释说明上述技术方案。
图3是根据本发明实施例的一种智能机器人的步态控制方法的流程示意图(一)；如图3所示：
S302：获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列，其中，动作序列包括但不限于指定时间到达指定地点完成指定动作；
S304：通过状态观测器获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息，所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息，所述外部环境的状态信息包括：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息，其中通过路径规划模块生成的运动路径信息，通过感知模块获取所述外部环境的图像信息；
S306：通过所述全身动力学模型得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解，所述关节电机的第四位置、第四速度和第四力矩；
S308：通过所述关节电机的第四位置、第四速度和第四力矩对所述智能机器人的关节进行控制。
图4是根据本发明实施例的一种智能机器人的步态控制方法的流程示意图(二)；
S402：获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列，其中，动作序列包括但不限于指定时间到达指定地点完成指定动作；
S404：通过状态观测器获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息，所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息，所述外部环境的状态信息包括：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息，其中通过路径规划模块生成的运动路径信息，通过感知模块获取所述外部环境的图像信息；
S406：通过所述单刚体模型得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解；
S408：通过所述全身动力学模型得到所述关节电机的第四位置、第四速度和第四力矩；
S410：通过所述关节电机的第四位置、第四速度和第四力矩对所述智能机器人的关节进行控制。
通过上述技术方案，由于可以在获取用户指令和获取状态信息的情况下，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，进而根据所述控制指令对所述智能机器人进行控制，从而使得智能机器人适应各种运动场景。因此，解决了现有技术中需要根据具体运动场景提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动场景的问题。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到根据上述实施例的方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如只读存储器(Read-Only Memory，简称为ROM)、随机存取存储器(Random Access Memory，简称为RAM)、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，或者网络设备等)执行本申请各个实施例所述的方法。
在本实施例中还提供了一种智能机器人的步态控制装置，该智能机器人的步态控制装置用于实现上述实施例及优选实施方式，已经进行过说明的不再赘述。如以下所使用的，术语“模块”可以实现预定功能的软件和/或硬件的组合。尽管以下实施例所描述的装置较佳地以软件来实现，但是硬件，或者软件和硬件的组合的实现也是可能并被构想的。
图5为根据本发明实施例的一种智能机器人的步态控制装置的结构框图，如图5所示：
第一获取模块50，用于获取用户指令，其中，用户指令用于控制至少以下之一：智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；
第二获取模块52，用于获取状态信息，其中，状态信息包括：智能机器人的状态信息和外部环境的状态信息；
接收模块54，用于将用户指令和状态信息输入智能机器人的全身动力学模型，得到智能机器人的控制指令；
控制模块56，用于根据控制指令对智能机器人进行控制。
通过上述技术方案，由于可以在获取用户指令和获取状态信息的情况下，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，进而根据所述控制指令对所述智能机器人进行控制，从而使得智能机器人适应各种运动场景。因此，解决了现有技术中需要根据具体运动场景提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动场景的问题。
需要说明的是，本发明实施例的智能机器人可以是双足智能机器人，三足智能机器人，还可以是四足智能机器人、六足智能机器人，本发明实施例对此不进行限定。
可选的，控制模块56还用于获取对所述智能机器人控制之后的控制结果信息；将所述控制结果信息添加到给所述状态信息的集合中。
需要说明的是，在根据所述控制指令对所述智能机器人进行控制之后，需要获取对所述智能机器人控制之后的控制结果信息，并将所述控制结果信息添加到给所述状态信息的集合中，以更新所述状态信息。在更新所述状态信息之后，循环执行步骤S202至S208，直到完成当前所述用户指令结束循环。
可选的，第二获取模块52还用于所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息；所述外部环境的状态信息包括：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息。
需要说明的是，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息。所述智能机器人的状态信息包括但不限于：所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息，其中，所述智能机器人的关节电机的第一位置、第一速度和第一力矩是由所述智能机器人的状态观测器测量得到的，所述智能机器人的位置姿势信息是惯性测量单元测量得到的。所述外部环境的状态信息包括：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息，其中，所述外部环境的图像信息是由所述智能机器人的感知模块获取的，感知模块可以是相机，红外摄像模组等用于拍摄的设备，所述智能机器人根据所述外部环境生成的运动路径信息，在实际操作中是所述智能机器人的路径规划模块生成的。
可选的，接收模块54还用于通过所述智能机器人的全身动力学模型，以所述用户指令的成功完成度为最优化目标，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以所述状态信息为参数，建立运动约束方程；根据最优化目标和所述运动约束方程，得 到所述智能机器人的控制指令。
需要说明的是，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，构建最优化问题，以所述用户指令的成功完成度为最优化目标，以所述状态信息为参数，所述智能机器人的全身动力学模型根据上述构建的最优化问题，对所述智能机器人的落脚点位置、脚底受力、落脚序列和时长进行求解，得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解。所述智能机器人的全身动力学模型再通过所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解计算所述智能机器人的第二位置、第二速度和第二力矩，进而得到所述智能机器人的控制指令。其中，所述智能机器人的控制指令包括所述关节电机的第四位置、第四速度和第四力矩。需要说明的是，所述第四位置、第四速度和第四力矩是所述智能机器人的全身动力学模型计算出来的，所述智能机器人的关节电机的第一位置、第一速度和第一力矩是由所述智能机器人的状态观测器测量得到的。需要说明的是，上述得到所述智能机器人的控制指令的方法都是在所述全身动力学模型中计算的，其中，由于计算量太大，上述计算方法很难以简单的约束条件求解，然后代入最优化目标来确定解，本发明计算上述问题采用的是QP求解的方法。通过上述技术手段，解决了现有技术存在需要根据具体运动情况提前定义智能机器人的运动序列，不能动态调整智能机器人的运动序列使得智能机器人适应各种运动情况的问题。
可选的，接收模块54中所述运动约束方程，包括：
等式约束1：M(q)ddq+H(q,dq)+G-J cF ext＝τ
等式约束2：P foot(t)＝a
a＝a1，t∈{t|t0＜t＜t1}；a＝a2，t∈{t|t1＜t＜t2}
不等式约束：F ext(t)＞b
b＝b1，t∈{t|t0＜t＜t1}；b＝b2，t∈{t|t1＜t＜t2}
其中，q代表所述智能机器人的关节的角度，dq代表所述关节的角速度，ddq代表所述关节的角加速度，G为重力，F ext和J c分别是所述智能机器人受到的外力和其对应的雅克比矩阵，M是惯性矩阵，H是向心力科氏力矩阵，M(q)是关于q的M函数，τ为所述智能机器人的关节控制力矩，P foot代表所述智能机器人足端位置，t0、t1、t2分别为与用户指定的动作序列所对应的时刻点，a和b是设置参数。
需要说明的是，所述运动约束方程是对智能机器人的行进路线的限制。举例说明，智能机器人从A地到达B地，有很多条路径，所述运动约束方程就可以理解是使用者对智能机器人从A地到达B地，比如只能向西走等约束条件。其中，上述运动约束包括但不限于智能机器 人脚底受力的摩擦锥约束、地形对智能机器人的机身位置姿势的限制。所述运动约束方程用到了雅克比矩阵，其中通过雅克比矩阵就可以把智能机器人受到的外力从笛卡尔空间转换到关节空间，进而，通过智能机器人的关节电机控制智能机器人行走。
可选的，接收模块54中所述最优化目标为min(X c-X t) TR(X c-X t)+τ TQτ，其中，X c和X t分别为所述智能机器人的当前状态信息和目标状态信息，X c和X t由所述智能机器人位置、速度、轨迹点所组成，τ为所述智能机器人的关节控制力矩，R和Q分别为权重矩阵。
需要说明的是，智能机器人获取的状态信息是当前状态信息，目标状态信息是所述智能机器人的控制指令中所述关节电机的位置、速度和力矩对应的所述智能机器人的状态信息。所述最优化目标是所述智能机器人的全身动力学模型根据所述运动约束方程求出的解的限制，比如，所述智能机器人的全身动力学模型根据所述运动约束方程求出了多个解，所述最优化目标可以是智能机器人从A地到达B地的多个路径中最短的路径，其中，本发明计算上述问题可以采用QP求解的方法。
可选的，接收模块54还用于通过所述智能机器人的单刚体模型，以所述用户指令的成功完成度为最优化目标，以所述状态信息为参数，对所述智能机器人的落脚点位置、脚底受力、落脚序列和时长进行求解，得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解；通过所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解和所述智能机器人的全身动力学模型计算所述智能机器人的关节电机的第二位置、第二速度和第二力矩，以得到所述智能机器人的控制指令。
需要说明的是，本实施例可以完全通过所述全身动力学模型得到所述智能机器人的控制指令，本发明还提供了一种通过单刚体模型和全身动力学模型得到所述智能机器人的控制指令的方法：
首先通过所述智能机器人的单刚体模型，以所述用户指令的成功完成度为最优化目标，以所述状态信息为参数，根据上述参数和目标，所述单刚体模型对所述智能机器人的落脚点位置、脚底受力、落脚序列和时长进行求解，得到所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解。然后将所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的解输入所述智能机器人的全身动力学模型，计算所述智能机器人的关节电机的第二位置、第二速度和第二力矩，进而得到所述智能机器人的控制指令。
可选的，接收模块54还用于检测所述用户指令是否发生变化：当检测到所述用户指令不变，循环执行以下步骤直到完成当前所述用户指令结束循环：获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令， 根据所述控制指令对所述智能机器人进行控制；当检测到所述用户指令发生变化，循环执行以下步骤直至完成当前所述用户指令结束循环：获取所述用户指令，获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。
需要说明的是，在根据所述控制指令对所述智能机器人进行控制之后，还需要检测所述用户指令是否发生变化。如果检测到所述用户指令不变，那么循环执行：获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。直到完成当前所述用户指令结束循环。如果检测到所述用户指令发生变化，循环执行：获取所述用户指令，获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。直至完成当前所述用户指令结束循环
需要说明的是，上述各个模块是可以通过软件或硬件来实现的，对于后者，可以通过以下方式实现，但不限于此：上述模块均位于同一处理器中；或者，上述各个模块以任意组合的形式分别位于不同的处理器中。
本发明的实施例还提供了一种计算机可读的存储介质，该存储介质中存储有计算机程序，其中，该计算机程序被设置为运行时执行上述任一项方法实施例中的步骤。
可选地，在本实施例中，上述存储介质可以被设置为存储用于执行以下步骤的计算机程序：S1，获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；
S2，获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息；
S3，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令；
S4，根据所述控制指令对所述智能机器人进行控制。
可选地，在本实施例中，上述存储介质可以包括但不限于：U盘、只读存储器ROM、随机存取存储器RAM、移动硬盘、磁碟或者光盘等各种可以存储计算机程序的介质。
本发明的实施例还提供了一种电子装置，包括存储器和处理器，该存储器中存储有计算机程序，该处理器被设置为运行计算机程序以执行上述任一项方法实施例中的步骤。
可选地，上述电子装置还可以包括传输设备以及输入输出设备，其中，该传输设备和上述处理器连接，该输入输出设备和上述处理器连接。
可选地，在本实施例中，上述处理器可以被设置为通过计算机程序执行以下步骤：
S1，获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；
S2，获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息；
S3，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令；
S4，根据所述控制指令对所述智能机器人进行控制。
可选地，本实施例中的具体示例可以参考上述实施例及可选实施方式中所描述的示例，本实施例在此不再赘述。
显然，本领域的技术人员应该明白，上述的本发明的各模块或各步骤可以用通用的计算装置来实现，它们可以集中在单个的计算装置上，或者分布在多个计算装置所组成的网络上，可选地，它们可以用计算装置可执行的程序代码来实现，从而，可以将它们存储在存储装置中由计算装置来执行，并且在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤，或者将它们分别制作成各个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样，本发明不限制于任何特定的硬件和软件结合。
以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (11)

  1. 一种智能机器人的步态控制方法，其特征在于，包括：
    获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人所要到达的目标位置、速度、轨迹点和动作序列；
    获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息；
    将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令；
    根据所述控制指令对所述智能机器人进行控制。
  2. 如权利要求1所述的智能机器人的步态控制方法，其特征在于，根据所述控制指令对所述智能机器人进行控制之后，所述方法还包括：
    获取对所述智能机器人控制之后的控制结果信息；
    将所述控制结果信息更新到所述状态信息的集合中。
  3. 如权利要求1所述的智能机器人的步态控制方法，其特征在于，所述智能机器人的状态信息包括至少以下之一：所述智能机器人的关节电机的第一位置、第一速度和第一力矩，以及所述智能机器人的位置姿势信息；
    所述外部环境的状态信息包括至少以下之一：所述外部环境的图像信息，以及所述智能机器人根据所述外部环境生成的运动路径信息。
  4. 如权利要求1所述的智能机器人的步态控制方法，其特征在于，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，包括：
    通过所述智能机器人的全身动力学模型，以所述用户指令的成功完成度为最优化目标，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以所述状态信息为参数，以全身动力学模型和足端位置、脚底受力为约束方程，建模求解最优化参数；
    根据最优化目标和所述运动约束方程，得到所述智能机器人的控制指令。
  5. 如权利要求4所述的智能机器人的步态控制方法，其特征在于，所述运动约束方程，包括：
    等式约束1：M(q)ddq+H(q,dq)+G-J cF ext＝τ
    等式约束2：P foot(t)＝a
    a＝a1，t∈{t|t0＜t＜t1}；a＝a2，t∈{t|t1＜t＜t2}
    不等式约束：F ext(t)＞b
    b＝b1，t∈{t|t0＜t＜t1}；b＝b2，t∈{t|t1＜t＜t2}
    其中，q代表所述智能机器人的关节的角度，dq代表所述关节的角速度，ddq代表所述关节的角加速度，G为重力，F ext和J c分别是所述智能机器人受到的外力和其对应的雅克比矩阵，M是惯性矩阵，H是向心力科氏力矩阵，M(q)是关于q的M函数，τ为所述智能机器人的关节控制力矩，P foot代表所述智能机器人足端位置，t0、t1、t2分别为与用户指定的动作序列所对应的时刻点，a和b是设置参数。
  6. 如权利要求4所述的智能机器人的步态控制方法，其特征在于，所述最优化目标为min(X c-X t) TR(X c-X t)+τ TQτ，其中，X c和X t分别为所述智能机器人的当前状态信息和目标状态信息，X c和X t由所述智能机器人位置、速度、轨迹点所组成，τ为所述智能机器人的关节控制力矩，R和Q分别为权重矩阵。
  7. 如权利要求1所述的智能机器人的步态控制方法，其特征在于，在根据所述控制指令对所述智能机器人进行控制之前，所述方法还包括：
    通过所述智能机器人的单刚体模型，以所述用户指令的成功完成度为最优化目标，该最优化目标以所述状态信息为参数，以所述智能机器人的落脚点位置、脚底受力、落脚序列和时长为变量，以单刚体运动学模型和足端位置、脚底受力为约束方程，建模求解最优化参数；
    通过所述智能机器人的落脚点位置、脚底受力、落脚序列和时长的最优解和所述智能机器人的全身动力学模型计算所述智能机器人的关节电机的第二位置、第二速度和第二力矩，以得到所述智能机器人的控制指令。
  8. 如权利要求1所述的智能机器人的步态控制方法，其特征在于，在根据所述控制指令对所述智能机器人进行控制之后，所述方法还包括：
    检测所述用户指令是否发生变化：
    当检测到所述用户指令不变，循环执行以下步骤直到完成当前所述用户指令结束循环：
    获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制；
    当检测到所述用户指令发生变化，循环执行以下步骤直至完成当前所述用户指令结束循环：
    获取所述用户指令，获取所述状态信息，将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令，根据所述控制指令对所述智能机器人进行控制。
  9. 一种智能机器人的步态控制方法装置，其特征在于，包括：
    第一获取模块，用于获取用户指令，其中，所述用户指令用于控制至少以下之一：所述智能机器人的所要到达的目标位置、速度、轨迹点和动作序列；
    第二获取模块，用于获取状态信息，其中，所述状态信息包括：所述智能机器人的状态信息和外部环境的状态信息；
    接收模块，用于将所述用户指令和所述状态信息输入所述智能机器人的全身动力学模型，得到所述智能机器人的控制指令；
    控制模块，用于根据所述控制指令对所述智能机器人进行控制。
  10. 一种计算机可读的存储介质，其特征在于，所述存储介质中存储有计算机程序，其中，所述计算机程序被设置为运行时执行所述权利要求1至7任一项中所述的方法。
  11. 一种电子装置，包括存储器和处理器，其特征在于，所述存储器中存储有计算机程序，所述处理器被设置为运行所述计算机程序以执行所述权利要求1至7任一项中所述的方法。
PCT/CN2022/091877 2021-05-18 2022-05-10 智能机器人的步态控制方法及装置、存储介质、电子装置 Ceased WO2022242501A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110537158.2A CN113033027B (zh) 2021-05-18 2021-05-18 智能机器人的步态控制方法及装置、存储介质、电子装置
CN202110537158.2 2021-05-18

Publications (1)

Publication Number Publication Date
WO2022242501A1 true WO2022242501A1 (zh) 2022-11-24

Family

ID=76455314

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/091877 Ceased WO2022242501A1 (zh) 2021-05-18 2022-05-10 智能机器人的步态控制方法及装置、存储介质、电子装置

Country Status (2)

Country Link
CN (2) CN113962073A (zh)
WO (1) WO2022242501A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113962073A (zh) * 2021-05-18 2022-01-21 追觅创新科技(苏州)有限公司 智能机器人的步态控制方法及装置、存储介质、电子装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109871018A (zh) * 2019-02-25 2019-06-11 北京航空航天大学 一种具有腰部自由度四足机器人的控制方法
CN112051741A (zh) * 2020-09-16 2020-12-08 北京理工大学 一种双足机器人动态运动生成与控制方法
CN112720462A (zh) * 2020-12-09 2021-04-30 深圳先进技术研究院 一种机器人的轨迹规划系统和方法
CN113031450A (zh) * 2021-05-18 2021-06-25 追创科技(苏州)有限公司 智能机器人的前馈控制方法及装置、存储介质、电子装置
CN113033027A (zh) * 2021-05-18 2021-06-25 追创科技(苏州)有限公司 智能机器人的步态控制方法及装置、存储介质、电子装置

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011067621A1 (en) * 2009-12-04 2011-06-09 Toyota Jidosha Kabushiki Kaisha Robot control system, motion data creation apparatus and its creating method
US10518412B2 (en) * 2015-05-13 2019-12-31 National Insitute of Advanced Industrial Science Robot behavior generation method
CN106547206B (zh) * 2016-06-29 2019-08-06 中南大学 一种基于脚力估计算法的多足机器人动态稳定性定量判定方法
WO2020133288A1 (zh) * 2018-12-28 2020-07-02 深圳市优必选科技有限公司 一种双足机器人步态控制方法以及双足机器人
CN111070201B (zh) * 2019-08-30 2022-09-16 南京理工大学 四足机器人在负载突变下基于zmp理论的反应式鲁棒控制方法
CN111208826B (zh) * 2020-02-18 2023-06-16 杭州云深处科技有限公司 四足机器人溜蹄步态规划方法、装置、设备及可读介质
CN112684794B (zh) * 2020-12-07 2022-12-20 杭州未名信科科技有限公司 基于元强化学习的足式机器人运动控制方法、装置及介质
CN112783175B (zh) * 2021-01-04 2022-09-16 腾讯科技（深圳）有限公司 质心轨迹确定方法、装置、足式机器人、设备及介质
CN112775976B (zh) * 2021-02-05 2022-05-10 深圳市优必选科技股份有限公司 任务执行控制方法、装置、控制设备及可读存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109871018A (zh) * 2019-02-25 2019-06-11 北京航空航天大学 一种具有腰部自由度四足机器人的控制方法
CN112051741A (zh) * 2020-09-16 2020-12-08 北京理工大学 一种双足机器人动态运动生成与控制方法
CN112720462A (zh) * 2020-12-09 2021-04-30 深圳先进技术研究院 一种机器人的轨迹规划系统和方法
CN113031450A (zh) * 2021-05-18 2021-06-25 追创科技(苏州)有限公司 智能机器人的前馈控制方法及装置、存储介质、电子装置
CN113033027A (zh) * 2021-05-18 2021-06-25 追创科技(苏州)有限公司 智能机器人的步态控制方法及装置、存储介质、电子装置
CN113962073A (zh) * 2021-05-18 2022-01-21 追觅创新科技(苏州)有限公司 智能机器人的步态控制方法及装置、存储介质、电子装置

Also Published As

Publication number Publication date
CN113962073A (zh) 2022-01-21
CN113033027B (zh) 2021-11-02
CN113033027A (zh) 2021-06-25

Similar Documents

Publication Publication Date Title
CN109483530B (zh) 一种基于深度强化学习的足式机器人运动控制方法及系统
JP6715952B2 (ja) バランスのとれた運動および行動能力が向上した移動型ロボット
CN112684794B (zh) 基于元强化学习的足式机器人运动控制方法、装置及介质
CN111506100A (zh) 多足机器人关节控制方法、装置和多足机器人
CN108058758B (zh) 基于六自由度腿的机器人双足行走仿真评估系统及评估方法
CN112051797A (zh) 足式机器人运动控制方法、装置、设备及介质
US12422849B2 (en) Method for building controller for robot, method, device for controlling motion of robot, and robot
CN116985110B (zh) 控制足式机器人的方法和装置及足式机器人
CN109813319A (zh) 一种基于slam建图的开环优化方法及系统
CN111421554B (zh) 基于边缘计算的机械臂智能控制系统、方法、装置
CN112847373B (zh) 一种机器人轨迹同步控制方法、计算机可读存储介质
CN116442245A (zh) 服务机器人的控制方法、装置、系统及存储介质
CN113500604A (zh) 一种机器人控制方法、装置、设备及存储介质
WO2023165174A1 (zh) 构建机器人的控制器的方法、机器人的运动控制方法、装置以及机器人
WO2022242501A1 (zh) 智能机器人的步态控制方法及装置、存储介质、电子装置
Chen et al. Multirobolearn: An open-source framework for multi-robot deep reinforcement learning
Mastrogeorgiou et al. Slope handling for quadruped robots using deep reinforcement learning and toe trajectory planning
CN108037661A (zh) 用于摄像机器人的运动轨迹设计方法
CN114723807B (zh) 质心轨迹确定方法、装置、足式机器人、设备及介质
CN105717925B (zh) 一种基于无线传感网和云计算的电脑鼠机器人控制系统
WO2023165192A1 (zh) 机器人控制方法、装置、机器人以及计算机可读存储介质
CN115781658A (zh) 构建机器人的控制器的方法和机器人
CN118131801A (zh) 运动控制方法及装置、运动轨迹生成方法及装置
CN113031450A (zh) 智能机器人的前馈控制方法及装置、存储介质、电子装置
EP4390594A1 (en) Moving trajectory determining method and apparatus, computer device, and storage medium

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22803823

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22803823

Country of ref document: EP

Kind code of ref document: A1